PREFACE

Business Research Methodology

This book is the result of my teaching experience in the subject BUSINESS RESEARCH METHOLDOGY and
working experience in RESEARCH & STATISTICS to Sikkim Manipal University,Udupi students for about 8
years.It is designed to meet the requirements of students at Bachelor’s, Masters and Ph D levels in
Engineering and Management(M E, MCA MBA and Ph D in Project Management,Engg,Computer
Applications and Business Administration).

The main highlight of the book is the theoretical and numerical solved problem approach framed by the
author with many theoretical concepts .This book has a large number of problems solved in some
chapters & many questions with answers and explanations.

I thank various International software makers in the field of Statistics which made me enable to work on
tricky NUMERICAL PROBLEMS involving data and almost all theoretical concepts r covered in this book
related to Business Research Methodology.

There are many problems and all theoretical concepts related to Statistics framed by myself and can be
best suitable for Bachelors,Masters & Ph D students during their RESEARCH WORK in the three fields
mentioned below:

ENGINEERING-ALL FIELDS.(BACHELOR LEVEL,MASTERS LEVEL AND DOCTORS LEVEL)

COMPUTER APPLICATIONS. (BACHELOR LEVEL,MASTERS LEVEL AND DOCTORS LEVEL)

BUSINESS ADMINISTRATION. (BACHELORLEVEL,MASTERS LEVEL AND DOCTORS LEVEL)

SRINIVAS R RAO

EDUNXT CERTIFIED LEVEL III FACULTY FOR MBA

TRACKS INDIA INFOTECH LTD,UDUPI

SIKKIM MANIPAL UNIVERSITY,MANIPAL

 ABOUT THE BOOK
This book is on BUSINESS RESEARCH METHODOLOGY

which is a compulsory subject for Commerce students .Even the higher level students and bachelor level
students can also read it as it contains a lot of numerical problems framed by me.

Chapter-I

Research – Qualities of Researcher – Components of Research Problem – Various Steps In

Scientific Research – Types of Research – Hypotheses Research Purposes - Research Design –
Survey Research – Case Study Research.

Chapter-II

Data Collection – Sources of Data – Primary Data – Secondary Data

- Procedure Questionnaire – Sampling Methods – Merits and Demerits – Experiments –

Observation Method – Sampling Errors - Type-I Error & Type-II Error.

Chapter-III

Statistical Analysis – Introduction To Statistics – Probability Theories – Conditional Probability,

Poisson Distribution, Binomial Distribution and Properties of Normal Distributions – Hypothesis
Tests

– One Sample Test – Two Sample Tests / Chi-Square Test, Association of Attributes - Standard
Deviation – Co-Efficient of Variations .

Chapter-IV

Statistical Applications – Correlation and Regression Analysis – Analysis of Variance – Partial

and Multiple Correlation – Factor Analysis and Conjoint Analysis – Multifactor Evaluation –
Two-Factor Evaluation Approaches.

Chapter-V

Research Reports – Structure and Components of Research Report

– Types of Report, Characteristics of Good Research Report, Pictures and Graphs, Introduction
To SPSS.
are the 5 chapters with various sub-topics covered in this book. All theories are given and explained and
all formulae are also covered in this book in Statistics section of the subject.

I feel that this is a unique book as there are theory,formulae & numerical problems solved with all
possible steps.

HAPPY READING.

THANKS

REGARDS

AUTHOR

(SRINIVAS R RAO)
ABOUT THE AUTHOR

Author’s name is Srinivas R Rao, born and done his school level education in Mangalore,Karnataka in a
reputed private school Canara High School and PUC(+2) from Canara PUC in Science stream with PCMB
as main subjects.
Later, pursuing LL.B(5 Years) course passed the degree in 1999 and done Diploma in Export
Management ,Diploma in Customs and Central Excise , Diploma in Business Administration and some
important IT subjects like MS-Office,Internet/Email,Visual Basic 6.0,C,C++,Java,Advanced Java,Oracle
with D2K,HTML with Javascript,VBscript and Active Server Pages.
Joined as a FACULTY for students in a small computer Institute in 2002 July and later after 4 months
worked in a company by name CRP Technologies(I) .P.Ltd as Branch Manager(Risk Manager) for
Mangalore,Udupi and Kasargod areas from January 26 2003 to June 11 2007.In the year 2005 pursued
MBA distance education course. Currently working as a FACULTY in Sikkim Manipal University , Udupi
centre for BBA & MBA students and teaching numerical subjects like Statistics/Operations Research(Mgt
Science/Quant. Techniques for Mgt)/Accounting and several numerical and difficult oriented subjects
for distance education students in their weekend contact classes from July 2010 till present day.
Thanks
Regards
Author
(SRINIVAS R RAO)
 BUSINESS RESEARCH METHODOLGY

TABLE OF CONTENTS:

Chapter-I

Research – Qualities of Researcher – Components of Research Problem – Various Steps In

Scientific Research – Types of Research – Hypotheses Research Purposes - Research Design
– Survey Research – Case Study Research.

Chapter-II

Data Collection – Sources of Data – Primary Data – Secondary Data

- Procedure Questionnaire – Sampling Methods – Merits and Demerits – Experiments –

Observation Method – Sampling Errors - Type-I Error & Type-II Error.

Chapter-III

Statistical Analysis – Introduction To Statistics – Probability Theories – Conditional

Probability, Poisson Distribution, Binomial Distribution and Properties of Normal
Distributions – Hypothesis Tests

– One Sample Test – Two Sample Tests / Chi-Square Test, Association of Attributes -
Standard Deviation – Co-Efficient of Variations .

Chapter-IV

Statistical Applications – Correlation and Regression Analysis – Analysis of Variance –

Partial and Multiple Correlation – Factor Analysis and Conjoint Analysis – Multifactor
Evaluation – Two-Factor Evaluation Approaches.

Chapter-V

Research Reports – Structure and Components of Research Report

– Types of Report, Characteristics of Good Research Report, Pictures and Graphs,

Introduction To SPSS.
 BUSINESS RESEARCH METHODOLOGY

Chapter-I
Research – Qualities of Researcher – Components of Research
Problem – Various Steps In Scientific Research – Types of Research –
Hypotheses Research Purposes - Research Design – Survey Research –
Case Study Research.

Chapter-II
Data Collection – Sources of Data – Primary Data – Secondary Data
- Procedure Questionnaire – Sampling Methods – Merits and Demerits –
Experiments – Observation Method – Sampling Errors - Type-I Error &
Type-II Error.

Chapter-III
Statistical Analysis – Introduction To Statistics – Probability
Theories – Conditional Probability, Poisson Distribution, Binomial
Distribution and Properties of Normal Distributions – Hypothesis Tests
– One Sample Test – Two Sample Tests / Chi-Square Test, Association of
Attributes - Standard Deviation – Co-Efficient of Variations .

Chapter-IV
Statistical Applications – Correlation and Regression Analysis –
Analysis of Variance – Partial and Multiple Correlation – Factor Analysis
and Conjoint Analysis – Multifactor Evaluation – Two-Factor Evaluation
Approaches.

1
Chapter-V
Research Reports – Structure and Components of Research Report
– Types of Report, Characteristics of Good Research Report, Pictures and
Graphs, Introduction To SPSS.

------

2
 CHAPTER – I

Introduction

1 Meaning, Objectives And Types Of Research

2 Qualities Of Researcher
3 Significance Of Research
4 Research Process
5 Research Problem
6 Features, Importance, Characteristics, Concepts And Types Of
Research Design
7 Case Study Research
8 Hypothesis And Its Testing
9 Sample Survey And Sampling Methods

1.1 Meaning Of Research:

Research in simple terms refers to search for knowledge. It is a

scientific and systematic search for information on a particular topic or
issue. It is also known as the art of scientific investigation. Several social
scientists have defined research in different ways.

In the Encyclopedia of Social Sciences, D. Slesinger and M.

Stephension (1930) defined research as “the manipulation of things,
concepts or symbols for the purpose of generalizing to extend, correct
or verify knowledge, whether that knowledge aids in the construction of
theory or in the practice of an art”.

According to Redman and Mory (1923), research is a “systematized

effort to gain new knowledge”. It is an academic activity and therefore the
term should be used in a technical sense. According to Clifford Woody
(kothari, 1988), research comprises “defining and redefining problems,
formulating hypotheses or suggested solutions; collecting, organizing

3
and evaluating data; making deductions and reaching conclusions; and
finally, carefully testing the conclusions to determine whether they fit the
formulated hypotheses”.

Thus, research is an original addition to the available knowledge,

which contributes to its further advancement. It is an attempt to pursue
truth through the methods of study, observation, comparison and
experiment. In sum, research is the search for knowledge, using objective
and systematic methods to find solution to a problem.

1 Objectives Of Research:

The objective of research is to find answers to the questions by

applying scientific procedures. In other words, the main aim of research
is to find out the truth which is hidden and has not yet been discovered.
Although every research study has its own specific objectives, the research
objectives may be broadly grouped as follows:

1. To gain familiarity with new insights into a phenomenon (i.e.,

formulative research studies);
2. To accurately portray the characteristics of a particular individual,
group, or a situation (i.e., descriptive research studies);
3. To analyse the frequency with which something occurs (i.e.,
diagnostic research studies); and
4. To examine the hypothesis of a causal relationship between two
variables (i.e., hypothesis-testing research studies).

2 Research Methods Versus Methodology:

Research methods include all those techniques/methods that are

adopted for conducting research. Thus, research techniques or methods
are the methods that the researchers adopt for conducting the research
studies.

on the other hand, research methodology is the way in which

research problems are solved systematically. It is a science of studying how
research is conducted scientifically. Under it, the researcher acquaints
himself/herself with the various steps generally adopted to study a
research problem, along with the underlying logic behind them. Hence, it
4
is not only important for the researcher to know the research techniques/
methods, but also the scientific approach called methodology.

3 Research Approaches:

There are two main approaches to research, namely quantitative

approach and qualitative approach. The quantitative approach involves
the collection of quantitative data, which are put to rigorous quantitative
analysis in a formal and rigid manner. This approach further includes
experimental, inferential, and simulation approaches to research.
Meanwhile, the qualitative approach uses the method of subjective
assessment of opinions, behaviour and attitudes. Research in such a
situation is a function of the researcher’s impressions and insights. The
results generated by this type of research are either in non-quantitative
form or in the form which cannot be put to rigorous quantitative analysis.
Usually, this approach uses techniques like indepth interviews, focus group
interviews, and projective techniques.

4 Types Of Research:

There are different types of research. The basic ones are as follows.

1. Descriptive Versus Analytical:

Descriptive research consists of surveys and fact-finding enquiries

of different types. The main objective of descriptive research is describing
the state of affairs as it prevails at the time of study. The term ‘ex post
facto research’ is quite often used for descriptive research studies in social
sciences and business research. The most distinguishing feature of this
method is that the researcher has no control over the variables here. He/she
has to only report what is happening or what has happened. Majority of
the ex post facto research projects are used for descriptive studies in which
the researcher attempts to examine phenomena, such as the consumers’
preferences, frequency of purchases, shopping, etc. Despite the inability
of the researchers to control the variables, ex post facto studies may also
comprise attempts by them to discover the causes of the selected problem.
The methods of research adopted in conducting descriptive research are
survey methods of all kinds, including correlational and comparative
methods.
5
 Meanwhile in the Analytical research, the researcher has to use the
already available facts or information, and analyse them to make a critical
evaluation of the subject.

2. Applied Versus Fundamental:

Research can also be applied or fundamental in nature. An attempt

to find a solution to an immediate problem encountered by a firm, an
industry, a business organisation, or the society is known as applied
research. Researchers engaged in such researches aim at drawing certain
conclusions confronting a concrete social or business problem.

On the other hand, fundamental research mainly concerns

generalizations and formulation of a theory. In other words, “Gathering
knowledge for knowledge’s sake is termed ‘pure’ or ‘basic’ research”
(Young in Kothari, 1988). Researches relating to pure mathematics or
concerning some natural phenomenon are instances of Fundamental
Research. Likewise, studies focusing on human behaviour also fall under
the category of fundamental research.

Thus, while the principal objective of applied research is to find a

solution to some pressing practical problem, the objective of basic research
is to find information with a broad base of application and add to the
already existing organized body of scientific knowledge.

3. Quantitative Versus Qualitative:

Quantitative research relates to aspects that can be quantified or

can be expressed in terms of quantity. It involves the measurement of
quantity or amount. Various available statistical and econometric methods
are adopted for analysis in such research. Which includes correlation,
regressions and time series analysis etc,.

On the other hand, Qualitative research is concerned with

qualitative phenomena, or more specifically, the aspects related to or
involving quality or kind. For example, an important type of qualitative
research is ‘Motivation Research’, which investigates into the reasons
for certain human behaviour. The main aim of this type of research is
discovering the underlying motives and desires of human beings by using
6
in-depth interviews. The other techniques employed in such research
are story completion tests, sentence completion tests, word association
tests, and other similar projective methods. Qualitative research is
particularly significant in the context of behavioural sciences, which aim
at discovering the underlying motives of human behaviour. Such research
helps to analyse the various factors that motivate human beings to behave
in a certain manner, besides contributing to an understanding of what
makes individuals like or dislike a particular thing. However, it is worth
noting that conducting qualitative research in practice is considerably a
difficult task. Hence, while undertaking such research, seeking guidance
from experienced expert researchers is important.

4. Conceptual Versus Empirical:

The research related to some abstract idea or theory is known as

Conceptual Research. Generally, philosophers and thinkers use it for
developing new concepts or for reinterpreting the existing ones. Empirical
Research, on the other hand, exclusively relies on the observation or
experience with hardly any regard for theory and system. Such research
is data based, which often comes up with conclusions that can be verified
through experiments or observation. Empirical research is also known as
experimental type of research, in which it is important to first collect the
facts and their sources, and actively take steps to stimulate the production
of desired information. In this type of research, the researcher first
formulates a working hypothesis, and then gathers sufficient facts to prove
or disprove the stated hypothesis. He/she formulates the experimental
design, which according to him/her would manipulate the variables, so as to
obtain the desired information. This type of research is thus characterized
by the researcher’s control over the variables under study. In simple term,
empirical research is most appropriate when an attempt is made to prove
that certain variables influence the other variables in some way. Therefore,
the results obtained by using the experimental or empirical studies are
considered to be the most powerful evidences for a given hypothesis.

5. Other Types Of Research:

The remaining types of research are variations of one or more of

the afore-mentioned type of research. They vary in terms of the purpose
of research, or the time required to complete it, or may be based on some
7
other similar factor. On the basis of time, research may either be in the
nature of one-time or longitudinal time series research. While the research
is restricted to a single time-period in the former case, it is conducted over
several time-periods in the latter case. Depending upon the environment
in which the research is to be conducted, it can also be laboratory research
or field-setting research, or simulation research, besides being diagnostic
or clinical in nature. Under such research, in-depth approaches or case
study method may be employed to analyse the basic causal relations.
These studies usually undertake a detailed in-depth analysis of the causes
of certain events of interest, and use very small samples and sharp data
collection methods. The research may also be explanatory in nature.
Formalized research studies consist of substantial structure and specific
hypotheses to be verified. As regards to historical research, sources like
historical documents, remains, etc. Are utilized to study past events or
ideas. It also includes philosophy of persons and groups of the past or any
remote point of time.

Research has also been classified into decision-oriented and

conclusion-oriented categories. The decision-oriented research is always
carried out as per the need of a decision maker and hence, the researcher has
no freedom to conduct the research according to his/her own desires. On
the other hand, in the case of Conclusion-oriented research, the researcher
is free to choose the problem, redesign the enquiry as it progresses and
even change conceptualization as he/she wishes to. Operations research is
a kind of decision-oriented research, where in scientific method is used in
providing the departments, a quantitative basis for decision-making with
respect to the activities under their purview.

5 Importance Of Knowing How To Conduct Research:

The importance of knowing how to conduct research are listed below:

i. The knowledge of research methodology provides training to new

researchers and enables them to do research properly. It helps them
to develop disciplined thinking or a ‘bent of mind’ to objectively
observe the field;

ii. The knowledge of doing research inculcates the ability to evaluate

and utilize the research findings with confidence;

8
 iii. The knowledge of research methodology equips the researcher with
the tools that help him/her to make the observations objectively;
and

iv. The knowledge of methodology helps the research consumers to

evaluate research and make rational decisions.

6 Qualities Of A Researcher:

It is important for a researcher to possess certain qualities to

conduct research. First and foremost, he being a scientist should be firmly
committed to the ‘articles of faith’ of the scientific methods of research.
This implies that a researcher should be a social science person in the
truest sense. Sir Michael Foster cited by (Wilkinson and Bhandarkar, 1979)
identified a few distinct qualities of a scientist. According to him, a true
research scientist should possess the following qualities:

(1) First of all, the nature of a researcher must be of the temperament

that vibrates in unison with the theme which he is searching. Hence, the
seeker of knowledge must be truthful with truthfulness of nature, which
is much more important, much more exacting than what is sometimes
known as truthfulness. The truthfulness relates to the desire for accuracy
of observation and precision of statement. Ensuring facts is the principle
rule of science, which is not an easy matter. The difficulty may arise due
to untrained eye, which fails to see anything beyond what it has the power
of seeing and sometimes even less than that. This may also be due to the
lack of discipline in the method of science. An unscientific individual
often remains satisfied with the expressions like approximately, almost,
or nearly, which is never what nature is. A real research cannot see two
things which differ, however minutely, as the same.

(2) A researcher must possess an alert mind. Nature is constantly

changing and revealing itself through various ways. A scientific researcher
must be keen and watchful to notice such changes, no matter how small or
insignificant they may appear. Such receptivity has to be cultivated slowly
and patiently over time by the researcher through practice. An individual
who is ignorant or not alert and receptive during his research will not
make a good researcher. He will fail as a good researcher if he has no keen
eyes or mind to observe the unusual changes behind the routine. Research
9
demands a systematic immersion into the subject matter by the researcher
grasp even the slightest hint that may culminate into significant research
problems. In this context, Cohen and Negal cited by (Selltiz et al, 1965;
Wilkinson and Bhandarkar, 1979) state that “the ability to perceive in some
brute experience the occasion of a problem is not a common talent among
men… it is a mark of scientific genius to be sensitive to difficulties where
less gifted people pass by untroubled by doubt”.

(3) Scientific enquiry is pre-eminently an intellectual effort. It

requires the moral quality of courage, which reflects the courage of a
steadfast endurance. The process of conducting research is not an easy
task. There are occasions when a research scientist might feel defeated or
completely lost. This is the stage when a researcher would need immense
courage and the sense of conviction. The researcher must learn the art of
enduring intellectual hardships. In the words of Darwin, “It’s dogged that
does it”.

In order to cultivate the afore-mentioned three qualities of a

researcher, a fourth one may be added. This is the quality of making
statements cautiously. According to Huxley, the assertion that outstrips the
evidence is not only a blunder but a crime (Thompson, 1975). A researcher
should cultivate the habit of reserving judgment when the required data
are insufficient.

7 Significance Of Research:

According to a famous Hudson Maxim, “All progress is born of

inquiry. Doubt is often better than overconfidence, for it leads to inquiry,
and inquiry leads to invention”. It brings out the significance of research,
increased amount of which makes the progress possible. Research
encourages scientific and inductive thinking, besides promoting the
development of logical habits of thinking and organisation. The role of
research in applied economics in the context of an economy or business
is greatly increasing in modern times. The increasingly complex nature
of government and business has raised the use of research in solving
operational problems. Research assumes significant role in the formulation
of economic policy for both, the government and business. It provides
the basis for almost all government policies of an economic system.
Government budget formulation, for example, depends particularly on the
10
analysis of needs and desires of people, and the availability of revenues,
which requires research. Research helps to formulate alternative policies,
in addition to examining the consequences of these alternatives. Thus,
research also facilitates the decision-making of policy-makers, although
in itself is not a part of research. In the process, research also helps in the
proper allocation of a country’s scarce resources.

Research is also necessary for collecting information on the social

and economic structure of an economy to understand the process of change
occurring in the country. Collection of statistical information, though
not a routine task, involves various research problems. Therefore, large
staff of research technicians or experts are engaged by the government
these days to undertake this work. Thus, research as a tool of government
economic policy formulation involves three distinct stages of operation:
(i) investigation of economic structure through continual compilation
of facts; (ii) diagnosis of events that are taking place and analysis of the
forces underlying them; and (iii) the prognosis i.e., the prediction of future
developments (Wilkinson and Bhandarkar, 1979).

Research also assumes significance in solving various operational

and planning problems associated with business and industry. In several
ways, operations research, market research and motivational research
are vital and their results assist in taking business decisions. Market
research refers to the investigation of the structure and development of
a market for the formulation of efficient policies relating to purchases,
production and sales. Operational research relates to the application
of logical, mathematical, and analytical techniques to find solution to
business problems, such as cost minimization or profit maximization, or
the optimization problems. Motivational research helps to determine why
people behave in the manner they do with respect to market characteristics.
More specifically, it is concerned with the analysis of the motivations
underlying consumer behaviour. All these researches are very useful for
business and industry, and are responsible for business decision-making.

Research is equally important to social scientists for analyzing

the social relationships and seeking explanations to various social
problems. It gives intellectual satisfaction of knowing things for the sake
of knowledge. It also possesses the practical utility for the social scientist
to gain knowledge so as to be able to do something better or in a more
11
efficient manner. The research in social sciences is concerned with both
knowledge for its own sake, and knowledge for what it can contribute to
solve practical problems.

2 Research Process:

Research process consists of a series of steps or actions required for

effectively conducting research. The following are the steps that provide
useful procedural guidelines regarding the conduct of research:

(1) Formulating the research problem;

(2) Extensive literature survey;
(3) Developing hypothesis;
(4) Preparing the research design;
(5) Determining sample design;
(6) Collecting data;
(7) Execution of the project;
(8) Analysis of data;
(9) Hypothesis testing;
(10) Generalization and interpretation, and
(11) Preparation of the report or presentation of the results.

In other words, it involves the formal write-up of conclusions.

3 Research Problem:

The first and foremost stage in the research process is to select and
properly define the research problem. A researcher should first identify
a problem and formulate it, so as to make it amenable or susceptible to
research. In general, a research problem refers to an unanswered question
that a researcher might encounter in the context of either a theoretical or
practical situation, which he/she would like to answer or find a solution
to. A research problem is generally said to exist if the following conditions
emerge (Kothari, 1988):

i. There should be an individual or an organisation, say X, to whom

the Problem can be attributed. The individual or the organization is
situated in an environment Y, which is governed by certain uncontrolled
variables Z;
12
ii. There should be at least two courses of action to be pursued, say A1
and A2. These courses of action are defined by one or more values of
the controlled variables. For example, the number of items purchased
at a specified time is said to be one course of action.

iii. There should be atleast two alternative possible outcomes of the said
courses of action, say B1 and B2. Of them, one alternative should be
preferable to the other. That is, atleast one outcome should be what the
researcher wants, which becomes an objective.

iv. The courses of possible action available must offer a chance to the
researcher to achieve the objective, but not the equal chance. Therefore,
if P(Bj / X, A, Y) represents the probability of the occurrence of an
outcome Bj when X selects Aj in Y, then P(B1 / X, A1,Y) ≠ P (B1 / X,
A2, Y). Putting it in simple words, it means that the choices must not
have equal efficiencies for the desired outcome.

Above all these conditions, the individual or organisation may be

said to have arrived at the research problem only if X does not know what
course of action to be taken is the best. In other words, X should have a
doubt about the solution. Thus, an individual or a group of persons can
be said to have a problem if they have more than one desired outcome.
They should have two or more alternative courses of action, which have
some but not equal efficiency. This is required for probing the desired
objectives, such that they have doubts about the best course of action to be
taken. Thus, the components of a research problem may be summarised
as:

i. There should be an individual or a group who have some difficulty

or problem.

ii. There should be some objective(s) to be pursued. A person or an

organization who wants nothing cannot have a problem.

iii. There should be alternative ways of pursuing the objective the

researcher wants to pursue. This implies that there should be more
than one alternative means available to the researcher. This is because
if the researcher has no choice of alternative means, he/she would
not have a problem.

13
 iv. There should be some doubt in the mind of the researcher about
the choice of alternative means. This implies that research should
answer the question relating to the relative efficiency or suitability of
the possible alternatives.

v. There should be a context to which the difficulty relates.

Thus, identification of a research problem is the pre-condition

to conducting research. A research problem is said to be the one which
requires a researcher to find the best available solution to the given problem.
That is, the researcher needs to find out the best course of action through
which the research objective may be achieved optimally in the context of
a given situation. Several factors may contribute to making the problem
complicated. For example, the environment may alter, thus affecting the
efficiencies of the alternative courses of action taken or the quality of the
outcomes. The number of alternative courses of action might be very large
and the individual not involved in making the decision may be affected by
the change in environment and may react to it favorably or unfavorably.
Other similar factors are also likely to cause such changes in the context
of research, all of which may be considered from the point of view of a
research problem.

4 Research Design:

The most important step after defining the research problem is

preparing the design of the research project, which is popularly known
as the ‘research design’. A research design helps to decide upon issues
like what, when, where, how much, by what means etc. With regard to
an enquiry or a research study. A research design is the arrangement
of conditions for collection and analysis of data in a manner that aims to
combine relevance to the research purpose with economy in procedure.
Infact, research design is the conceptual structure within which research
is conducted; it constitutes the blueprint for the collection, measurement
and analysis of data (Selltiz et al, 1962). Thus, research design provides
an outline of what the researcher is going to do in terms of framing
the hypothesis, its operational implications and the final data analysis.
Specifically, the research design highlights decisions which include:

14
 1. The nature of the study
2. The purpose of the study
3. The location where the study would be conducted
4. The nature of data required
5. From where the required data can be collected
6. What time period the study would cover
7. The type of sample design that would be used
8. The techniques of data collection that would be used
9. The methods of data analysis that would be adopted and
10. The manner in which the report would be prepared

In view of the stated research design decisions, the overall research design
may be divided into the following (Kothari 1988):

a. The sampling design that deals with the method of selecting items
to be observed for the selected study;
b. The observational design that relates to the conditions under which
the observations are to be made;
c. The statistical design that concerns with the question of how many
items are to be observed, and how the information and data gathered
are to be analysed; and
d. The operational design that deals with the techniques by which the
procedures specified in the sampling, statistical and observational
designs can be carried out.

4.1 Features Of Research Design:

The important features of Research Design may be outlined as follows:

i. It constitutes a plan that identifies the types and sources of

information required for the research problem;
ii. It constitutes a strategy that specifies the methods of data collection
and analysis which would be adopted; and
iii. It also specifies the time period of research and monetary budget
involved in conducting the study, which comprise the two major
constraints of undertaking any research

15
4.2 Concepts Relating To Research Design:

Some of the important concepts relating to Research Design are

discussed below:

1. Dependent And Independent Variables:

A magnitude that varies is known as a variable. The concept

may assume different quantitative values like height, weight, income etc.
Qualitative variables are not quantifiable in the strictest sense of the term.
However, the qualitative phenomena may also be quantified in terms of the
presence or absence of the attribute(s) considered. The phenomena that
assume different values quantitatively even in decimal points are known
as ‘continuous variables’. But all variables need not be continuous. Values
that can be expressed only in integer values are called ‘non-continuous
variables’. In statistical terms, they are also known as ‘discrete variables’.
For example, age is a continuous variable, whereas the number of children
is a non-continuous variable. When changes in one variable depend upon
the changes in other variable or variables, it is known as a dependent
or endogenous variable, and the variables that cause the changes in the
dependent variable are known as the independent or explanatory or
exogenous variables. For example, if demand depends upon price, then
demand is a dependent variable, while price is the independent variable.
And, if more variables determine demand, like income and price of the
substitute commodity, then demand also depends upon them in addition
to the price of original commodity. In other words, demand is a dependent
variable which is determined by the independent variables like price of the
original commodity, income and price of substitutes.

2. Extraneous Variables:

The independent variables which are not directly related to the

purpose of the study but affect the dependent variables, are known as
extraneous variables. For instance, assume that a researcher wants to
test the hypothesis that there is a relationship between children’s school
performance and their self-confidence, in which case the latter is an
independent variable and the former, a dependent variable. In this context,
intelligence may also influence the school performance. However, since
it is not directly related to the purpose of the study undertaken by the
16
researcher, it would be known as an extraneous variable. The influence
caused by the extraneous variable(s) on the dependent variable is
technically called the ‘experimental error’. Therefore, a research study
should always be framed in such a manner that the influence of extraneous
variables on the dependent variable/s is completely controlled, and the
influence of independent variable/s is clearly evident.

3. Control:

One of the most important features of a good research design is

to minimize the effect of extraneous variable(s). Technically, the term
‘control’ is used when a researcher designs the study in such a manner that
it minimizes the effects of extraneous variables. The term ‘control’ is used
in experimental research to reflect the restrain in experimental conditions.

4. Confounded Relationship:

The relationship between the dependent and independent variables

is said to be confounded by an extraneous variable, when the dependent
variable is not free from its effects.

5. Research Hypothesis:

When a prediction or a hypothesized relationship is tested by

adopting scientific methods, it is known as research hypothesis. The
research hypothesis is a predictive statement which relates to a dependent
variable and an independent variable. Generally, a research hypothesis
must consist of at least one dependent variable and one independent
variable. Whereas, the relationships that are assumed but not to be tested
are predictive statements that are not to be objectively verified, thus are
not classified as research hypotheses.

6. Experimental and Non-experimental Hypothesis Testing Research:

When the objective of a research is to test a research hypothesis,

it is known as hypothesis-testing research. Such research may be in the
nature of experimental design or non-experimental design. The research in
which the independent variable is manipulated is known as ‘experimental
hypothesis-testing research’, whereas the research in which the independent
17
variable is not manipulated is termed as ‘non-experimental hypothesis-
testing research’. For example, assume that a researcher wants to examine
whether family income influences the school attendance of a group of
students, by calculating the coefficient of correlation between the two
variables. Such an example is known as a non-experimental hypothesis-
testing research, because the independent variable - family income is not
manipulated here. Again assume that the researcher randomly selects
150 students from a group of students who pay their school fees regularly
and then classifies them into two sub-groups by randomly including 75 in
Group A, whose parents have regular earning, and 75 in Group B, whose
parents do not have regular earning. Assume that at the end of the study,
the researcher conducts a test on each group in order to examine the effects
of regular earnings of the parents on the school attendance of the student.
Such a study is an example of experimental hypothesis-testing research,
because in this particular study the independent variable regular earnings
of the parents have been manipulated.

7. Experimental And Control Groups:

When a group is exposed to usual conditions in an experimental

hypothesis-testing research, it is known as ‘control group’. On the other
hand, when the group is exposed to certain new or special condition, it is
known as an ‘experimental group’. In the afore-mentioned example, Group
A can be called as control group and Group B as experimental group. If
both the groups, A and B are exposed to some special feature, then both
the groups may be called as ‘experimental groups’. A research design may
include only the experimental group or both the experimental and control
groups together.

8. Treatments:

Treatments refer to the different conditions to which the

experimental and control groups are subject to. In the example considered,
the two treatments are the parents with regular earnings and those with
no regular earnings. Likewise, if a research study attempts to examine
through an experiment the comparative effect of three different types of
fertilizers on the yield of rice crop, then the three types of fertilizers would
be treated as the three treatments.

18
9. Experiment:

Experiment refers to the process of verifying the truth of a

statistical hypothesis relating to a given research problem. For instance,
an experiment may be conducted to examine the yield of a certain new
variety of rice crop developed. Further, Experiments may be categorized
into two types, namely, ‘absolute experiment’ and ‘comparative experiment’.
If a researcher wishes to determine the impact of a chemical fertilizer on
the yield of a particular variety of rice crop, then it is known as absolute
experiment. Meanwhile, if the researcher wishes to determine the impact
of chemical fertilizer as compared to the impact of bio-fertilizer, then the
experiment is known as a comparative experiment.

10. Experimental Unit(s):

Experimental units refer to the pre-determined plots, characteristics

or the blocks, to which different treatments are applied. It is worth
mentioning here that such experimental units must be selected with great
caution.

4.3 Types Of Research Design:

There are different types of research designs. They may be broadly

categorized as:

(1) Exploratory Research Design;

(2) Descriptive and Diagnostic Research Design; and
(3) Hypothesis-Testing Research Design.

1. Exploratory Research Design:

The Exploratory Research Design is known as formulative research

design. The main objective of using such a research design is to formulate
a research problem for an in-depth or more precise investigation, or for
developing a working hypothesis from an operational aspect. The major
purpose of such studies is the discovery of ideas and insights. Therefore,
such a research design suitable for such a study should be flexible enough to
provide opportunity for considering different dimensions of the problem
under study. The in-built flexibility in research design is required as the
19
initial research problem would be transformed into a more precise one
in the exploratory study, which in turn may necessitate changes in the
research procedure for collecting relevant data. Usually, the following
three methods are considered in the context of a research design for such
studies. They are (a) a survey of related literature; (b) experience survey;
and (c) analysis of ‘insight-stimulating’ instances.

2. Descriptive And Diagnostic Research Design:

A Descriptive Research Design is concerned with describing

the characteristics of a particular individual or a group. Meanwhile, a
diagnostic research design determines the frequency with which a variable
occurs or its relationship with another variable. In other words, the study
analyzing whether a certain variable is associated with another comprises
a diagnostic research study. On the other hand, a study that is concerned
with specific predictions or with the narration of facts and characteristics
related to an individual, group or situation, are instances of descriptive
research studies. Generally, most of the social research design falls under
this category. As a research design, both the descriptive and diagnostic
studies share common requirements, hence they are grouped together.
However, the procedure to be used and the research design need to planned
carefully. The research design must also make appropriate provision for
protection against bias and thus maximize reliability, with due regard
to the completion of the research study in an economical manner. The
research design in such studies should be rigid and not flexible. Besides, it
must also focus attention on the following:

a) Formulation of the objectives of the study,

b) Proper designing of the methods of data collection,
c) Sample selection,
d) Data collection,
e) Processing and analysis of the collected data, and
f) Reporting the findings.

20
3. Hypothesis-Testing Research Design:

Hypothesis-Testing Research Designs are those in which the

researcher tests the hypothesis of causal relationship between two or more
variables. These studies require procedures that would not only decrease
bias and enhance reliability, but also facilitate deriving inferences about
the causality. Generally, experiments satisfy such requirements. Hence,
when research design is discussed in such studies, it often refers to the
design of experiments.

4.4 Importance Of Research Design:

The need for a research design arises out of the fact that it facilitates
the smooth conduct of the various stages of research. It contributes to
making research as efficient as possible, thus yielding the maximum
information with minimum effort, time and expenditure. A research
design helps to plan in advance, the methods to be employed for collecting
the relevant data and the techniques to be adopted for their analysis. This
would help in pursuing the objectives of the research in the best possible
manner, provided the available staff, time and money are given. Hence,
the research design should be prepared with utmost care, so as to avoid
any error that may disturb the entire project. Thus, research design plays a
crucial role in attaining the reliability of the results obtained, which forms
the strong foundation of the entire process of the research work.

Despite its significance, the purpose of a well-planned design is

not realized at times. This is because it is not given the importance that it
deserves. As a consequence, many researchers are not able to achieve the
purpose for which the research designs are formulated, due to which they
end up arriving at misleading conclusions. Therefore, faulty designing of
the research project tends to render the research exercise meaningless.
This makes it imperative that an efficient and suitable research design
must be planned before commencing the process of research. The research
design helps the researcher to organize his/her ideas in a proper form,
which in turn facilitates him/her to identify the inadequacies and faults in
them. The research design is also discussed with other experts for their
comments and critical evaluation, without which it would be difficult
for any critic to provide a comprehensive review and comments on the
proposed study.
21
4.5 Characteristics Of A Good Research Design:

A good research design often possesses the qualities of being flexible,

suitable, efficient, economical and so on. Generally, a research design which
minimizes bias and maximizes the reliability of the data collected and
analysed is considered a good design (Kothari 1988). A research design
which does not allow even the smallest experimental error is said to be
the best design for investigation. Further, a research design that yields
maximum information and provides an opportunity of viewing the various
dimensions of a research problem is considered to be the most appropriate
and efficient design. Thus, the question of a good design relates to the
purpose or objective and nature of the research problem studied. While a
research design may be good, it may not be equally suitable to all studies.
In other words, it may be lacking in one aspect or the other in the case of
some other research problems. Therefore, no single research design can
be applied to all types of research problems.

A research design suitable for a specific research problem would usually

involve the following considerations:

i. The methods of gathering the information;

ii. The skills and availability of the researcher and his/her staff, if
any;
iii. The objectives of the research problem being studied;
iv. The nature of the research problem being studied; and
v. The available monetary support and duration of time for the
research work.

5 Case Study Research:

The method of exploring and analyzing the life or functioning of

a social or economic unit, such as a person, a family, a community, an
institution, a firm or an industry is called case study method. The objective
of case study method is to examine the factors that cause the behavioural
patterns of a given unit and its relationship with the environment. The
data for a study are always gathered with the purpose of tracing the natural
history of a social or economic unit, and its relationship with the social or
economic factors, besides the forces involved in its environment. Thus,
a researcher conducting a study using the case study method attempts to
22
understand the complexity of factors that are operative within a social or
economic unit as an integrated totality. Burgess (Kothari, 1988) described
the special significance of the case study in understanding the complex
behaviour and situations in specific detail. In the context of social research,
he called such data as social microscope.

5.1 Criteria For Evaluating Adequacy Of Case Study:

John Dollard (Dollard, 1935) specified seven criteria for evaluating

the adequacy of a case or life history in the context of social research. They
are:

i. The subject being studied must be viewed as a specimen in a cultural

set up. That is, the case selected from its total context for the purpose of
study should be considered a member of the particular cultural group
or community. The scrutiny of the life history of the individual must
be carried out with a view to identify the community values, standards
and shared ways of life.

ii. The organic motors of action should be socially relevant. This is to say
that the action of the individual cases should be viewed as a series of
reactions to social stimuli or situations. To put in simple words, the
social meaning of behaviour should be taken into consideration.

iii. The crucial role of the family-group in transmitting the culture should
be recognized. This means, as an individual is the member of a family,
the role of the family in shaping his/her behaviour should never be
ignored.

iv. The specific method of conversion of organic material into social

behaviour should be clearly demonstrated. For instance, case-histories
that discuss in detail how basically a biological organism, that is man,
gradually transforms into a social person are particularly important.

v. The constant transformation of character of experience from

childhood to adulthood should be emphasized. That is, the life-
history should portray the inter-relationship between the individual’s
various experiences during his/her life span. Such a study provides a
comprehensive understanding of an individual’s life as a continuum.

23
vi. The ‘social situation’ that contributed to the individual’s gradual
transformation should carefully and continuously be specified as a
factor. One of the crucial criteria for life-history is that an individual’s
life should be depicted as evolving itself in the context of a specific
social situation and partially caused by it.

vii. The life-history details themselves should be organized according

to some conceptual framework, which in turn would facilitate their
generalizations at higher levels.

These criteria discussed by Dollard emphasize the specific link of

co-ordinated, related, continuous and configured experience in a cultural
pattern that motivated the social and personal behaviour. Although, the
criteria indicated by Dollard are principally perfect, some of them are
difficult to put to practice.

Dollard (1935) attempted to express the diverse events depicted in
the life-histories of persons during the course of repeated interviews by
utilizing psycho-analytical techniques in a given situational context. His
criteria of life-history originated directly from this experience. While the
life-histories possess independent significance as research documents, the
interviews recorded by the investigators can afford, as Dollard observed,
“rich insights into the nature of the social situations experienced by them”.

It is a well-known fact that an individual’s life is very complex.
Till date there is hardly any technique that can establish some kind of
uniformity, and as a result ensure the cumulative of case-history materials
by isolating the complex totality of a human life. Nevertheless, although
case history data are difficult to put to rigorous analysis, a skilful handling
and interpretation of such data could help in developing insights into
cultural conflicts and problems arising out of cultural-change.

Gordon Allport in (Kothari 1988) has recommended the following

aspects so as to broaden the perspective of case-study data:

i. If the life-history is written in first person, it should be as

comprehensive and coherent as possible.
ii. Life-histories must be written for knowledgeable persons. tThat is, if
the enquiry of study is sociological in nature, the researcher should

24
 write it on the assumption that it would be read largely by sociologists
only.
iii. It would be advisable to supplement case study data by observational,
statistical and historical data, as they provide standards for assessing
the reliability and consistency of the case study materials. Further,
such data offer a basis for generalizations.
iv. Efforts must be made to verify the reliability of life-history data by
examining the internal consistency of the collected material, and by
repeating the interviews with the concerned person. Besides this,
personal interviews with the persons who are well-acquainted with
him/her, belonging to his/her own group should be conducted.
v. A judicious combination of different techniques for data-collection
is crucial for collecting data that are culturally meaningful and
scientifically significant.
vi. Life-histories or case-histories may be considered as an adequate basis
for generalization to the extent that they are typical or representative
of a certain group.
vii. The researcher engaged in the collection of case study data should
never ignore the unique or typical cases. He/she should include them
as exceptional cases.

Case histories are filled with valuable information of a personal
or private nature. Such information not only helps the researcher to
portray the personality of the individual, but also the social background
that contributed to it. Besides, it also helps in the formulation of relevant
hypotheses. In general, although Blummer (in Wilkinson and Bhandarkar,
1979) was critical of documentary material, he gave due credit to case
histories by acknowledging the fact that the personal documents offer an
opportunity to the researcher to develop his/her spirit of enquiry. The
analysis of a particular subject would be more effective if the researcher
acquires close acquaintance with it through personal documents. However,
Blummer also acknowledges the limitations of the personal documents.
According to him, such documents do not entirely fulfill the criteria of
adequacy, reliability, and representativeness. Despite these shortcomings,
avoiding their use in any scientific study of personal life would be wrong,
as these documents become necessary and significant for both theory-
building and practice.

25
 In spite of these formidable limitations, case study data are used
by anthropologists, sociologists, economists and industrial psychiatrists.
Gordon Allport (Kothari, 1988) strongly recommends the use of case study
data for in-depth analysis of a subject. For, it is one’s acquaintance with
an individual that instills a desire to know his/her nature and understand
them. The first stage involves understanding the individual and all the
complexity of his/her nature. Any haste in analyzing and classifying the
individual would create the risk of reducing his/her emotional world into
artificial bits. As a consequence, the important emotional organizations,
anchorages and natural identifications characterizing the personal life
of the individual might not yield adequate representation. Hence, the
researcher should understand the life of the subject. Therefore, the totality
of life-processes reflected in the well-ordered life-history documents
become invaluable source of stimulating insights. Such life-history
documents provide the basis for comparisons that contribute to statistical
generalizations and help to draw inferences regarding the uniformities
in human behaviour, which are of great value. Even if some personal
documents do not provide ordered data about personal lives of people,
which is the basis of psychological science, they should not be ignored.
This is because the final aim of science is to understand, control and make
predictions about human life. Once they are satisfied, the theoretical
and practical importance of personal documents must be recognized as
significant. Thus, a case study may be considered as the beginning and the
final destination of abstract knowledge.

6 Hypothesis:

“Hypothesis may be defined as a proposition or a set of propositions

set forth as an explanation for the occurrence of some specified group
of phenomena either asserted merely as a provisional conjecture to guide
some investigation in the light of established facts” (Kothari, 1988). A
research hypothesis is quite often a predictive statement, which is capable
of being tested using scientific methods that involve an independent and
some dependent variables. For instance, the following statements may be
considered:

i. “Students who take tuitions perform better than the others who do
not receive tuitions” or,
ii. “The female students perform as well as the male students”.
26
 These two statements are hypotheses that can be objectively verified
and tested. Thus, they indicate that a hypothesis states what one is looking
for. Besides, it is a proposition that can be put to test in order to examine
its validity.

6.1 Characteristics Of Hypothesis:

A hypothesis should have the following characteristic features:-

i. A hypothesis must be precise and clear. If it is not precise and clear,

then the inferences drawn on its basis would not be reliable.

ii. A hypothesis must be capable of being put to test. Quite often, the
research programmes fail owing to its incapability of being subject to
testing for validity. Therefore, some prior study may be conducted by
the researcher in order to make a hypothesis testable. A hypothesis
“is tested if other deductions can be made from it, which in turn can
be confirmed or disproved by observation” (Kothari, 1988).

iii. A hypothesis must state relationship between two variables, in the

case of relational hypotheses.

iv. A hypothesis must be specific and limited in scope. This is because a

simpler hypothesis generally would be easier to test for the researcher.
And therefore, he/she must formulate such hypotheses.

v. As far as possible, a hypothesis must be stated in the simplest language,

so as to make it understood by all concerned. However, it should be
noted that simplicity of a hypothesis is not related to its significance.

vi. A hypothesis must be consistent and derived from the most known
facts. In other words, it should be consistent with a substantial body
of established facts. That is, it must be in the form of a statement
which is most likely to occur.

vii. A hypothesis must be amenable to testing within a stipulated or

reasonable period of time. No matter how excellent a hypothesis,
a researcher should not use it if it cannot be tested within a given
period of time, as no one can afford to spend a life-time on collecting
data to test it.

27
viii. A hypothesis should state the facts that give rise to the necessity of
looking for an explanation. This is to say that by using the hypothesis,
and other known and accepted generalizations, a researcher must be
able to derive the original problem condition. Therefore, a hypothesis
should explain what it actually wants to explain, and for this it should
also have an empirical reference.

6.2 Concepts Relating To Testing Of Hypotheses:

Testing of hypotheses requires a researcher to be familiar with

various concepts concerned with it such as:

1) Null Hypothesis And Alternative Hypothesis:

In the context of statistical analysis, hypotheses are of two types

viz., null hypothesis and alternative hypothesis. When two methods A
and B are compared on their relative superiority, and it is assumed that
both the methods are equally good, then such a statement is called as the
null hypothesis. On the other hand, if method A is considered relatively
superior to method B, or vice-versa, then such a statement is known as
an alternative hypothesis. The null hypothesis is expressed as H0, while
the alternative hypothesis is expressed as Ha. For example, if a researcher
wants to test the hypothesis that the population mean (μ) is equal to the
hypothesized mean (H0) = 100, then the null hypothesis should be stated as
the population mean is equal to the hypothesized mean 100. Symbolically
it may be written as:-

H 0: = μ = μ H0 = 100

If sample results do not support this null hypothesis, then it should

be concluded that something else is true. The conclusion of rejecting the
null hypothesis is called as alternative hypothesis H1. To put it in simple
words, the set of alternatives to the null hypothesis is termed as the
alternative hypothesis. If H0 is accepted, then it implies that Ha is being
rejected. On the other hand, if H0 is rejected, it means that Ha is being
accepted. For H0: μ = μ H0 = 100, the following three possible alternative
hypotheses may be considered:

28
 Alternative hypothesis To be read as follows
The alternative hypothesis is that the
H 1: μ ≠ μ H 0
population mean is not equal to 100, i.e., it
could be greater than or less than 100
The alternative hypothesis is that the
H1 : μ > μ H0
population mean is greater than 100
The alternative hypothesis is that the
H1 : μ < μ H0
population mean is less than 100

Before the sample is drawn, the researcher has to state the null
hypothesis and the alternative hypothesis. While formulating the null
hypothesis, the following aspects need to be considered:

A. Alternative hypothesis is usually the one which a researcher wishes

to prove, whereas the null hypothesis is the one which he/she wishes
to disprove. Thus, a null hypothesis is usually the one which a
researcher tries to reject, while an alternative hypothesis is the one
that represents all other possibilities.

B. The rejection of a hypothesis when it is actually true involves

great risk, as it indicates that it is a null hypothesis because then
the probability of rejecting it when it is true is α (i.e., the level of
significance) which is chosen very small.

C. Null hypothesis should always be specific hypothesis i.e., it should

not state about or approximately a certain value.

2) The Level Of Significance:

In the context of hypothesis testing, the level of significance is a very

important concept. It is a certain percentage that should be chosen with
great care, reason and insight. If for instance, the significance level is taken
at 5 per cent, then it means that H0 would be rejected when the sampling
result has a less than 0.05 probability of occurrence when H0 is true. In other
words, the five per cent level of significance implies that the researcher is
willing to take a risk of five per cent of rejecting the null hypothesis, when
(H0) is actually true. In sum, the significance level reflects the maximum
value of the probability of rejecting H0 when it is actually true, and which
is usually determined prior to testing the hypothesis.
29
3) Test Of Hypothesis Or Decision Rule:

Suppose the given hypothesis is H0 and the alternative hypothesis

H1, then the researcher has to make a rule known as the decision rule.
According to the decision rule, the researcher accepts or rejects H0. For
example, if the H0 is that certain students are good against the H1 that all
the students are good, then the researcher should decide the number of
items to be tested and the criteria on the basis of which to accept or reject
the hypothesis.

4) Type I And Type II Errors:

As regards the testing of hypotheses, a researcher can make basically

two types of errors. He/she may reject H0 when it is true, or accept H0
when it is not true. The former is called as Type I error and the latter is
known as Type II error. In other words, Type I error implies the rejection of
a hypothesis when it must have been accepted, while Type II error implies
the acceptance of a hypothesis which must have been rejected. Type I
error is denoted by α (alpha) and is known as α error, while Type II error
is usually denoted by β (beta) and is known as β error.

5) One-Tailed And Two-Tailed Tests:

These two types of tests are very important in the context of

hypothesis testing. A two-tailed test rejects the null hypothesis, when
the sample mean is significantly greater or lower than the hypothesized
value of the mean of the population. Such a test is suitable when the null
hypothesis is some specified value, the alternative hypothesis is a value
that is not equal to the specified value of the null hypothesis.

6.3 Procedure Of Hypothesis Testing:

Testing a hypothesis refers to verifying whether the hypothesis is

valid or not. Hypothesis testing attempts to check whether to accept or not
to accept the null hypothesis. The procedure of hypothesis testing includes
all the steps that a researcher undertakes for making a choice between the
two alternative actions of rejecting or accepting a null hypothesis. The
various steps involved in hypothesis testing are as follows:

30
1) Making a Formal Statement:

This step involves making a formal statement of the null hypothesis

(H0) and the alternative hypothesis (Ha). This implies that the hypotheses
should be clearly stated within the purview of the research problem. For
example, suppose a school teacher wants to test the understanding capacity
of the students which must be rated more than 90 per cent in terms of
marks, the hypotheses may be stated as follows:

Null Hypothesis H0 : = 100 Alternative Hypothesis H1 : > 100

2) Selecting A Significance Level:

The hypotheses should be tested on a pre-determined level of

significance, which should be specified. Usually, either 5% level or 1%
level is considered for the purpose. The factors that determine the levels
of significance are: (a) the magnitude of difference between the sample
means; (b) the sample size: (c) the variability of measurements within
samples; and (d) whether the hypothesis is directional or non-directional
(Kothari, 1988). In sum, the level of significance should be sufficient in
the context of the nature and purpose of enquiry.

3) Deciding The Distribution To Use:

After making decision on the level of significance for hypothesis

testing, the researcher has to next determine the appropriate sampling
distribution. The choice to be made generally relates to normal distribution
and the t-distribution. The rules governing the selection of the correct
distribution are similar to the ones already discussed with respect to
estimation.

4) Selection Of A Random Sample And Computing An Appropriate

Value:

Another step involved in hypothesis testing is the selection of a

random sample and then computing a suitable value from the sample data
relating to test statistic by using the appropriate distribution. In other
words, it involves drawing a sample for furnishing empirical data.

31
5) Calculation Of The Probability:

The next step for the researcher is to calculate the probability that
the sample result would diverge as far as it can from expectations, under
the situation when the null hypothesis is actually true.

6) Comparing The Probability:

Another step involved consists of making a comparison of the

probability calculated with the specified value of α, i.e. The significance
level. If the calculated probability works out to be equal to or smaller
than the α value in case of one-tailed test, then the null hypothesis is to be
rejected. On the other hand, if the calculated probability is greater, then
the null hypothesis is to be accepted. In case the null hypothesis H0 is
rejected, the researcher runs the risk of committing the Type I error. But,
if the null hypothesis H0 is accepted, then it involves some risk (which
cannot be specified in size as long as H0 is vague and not specific) of
committing the Type II error.

7 Sample Survey:

A sample design is a definite plan for obtaining a sample from a

given population (Kothari, 1988). Sample constitutes a certain portion of
the population or universe. Sampling design refers to the technique or the
procedure the researcher adopts for selecting items for the sample from
the population or universe. A sample design helps to decide the number
of items to be included in the sample, i.e., the size of the sample. The
sample design should be determined prior to data collection. There are
different kinds of sample designs which a researcher can choose. Some
of them are relatively more precise and easier to adopt than the others. A
researcher should prepare or select a sample design, which must be reliable
and suitable for the research study proposed to be undertaken.

8.1 Steps In Sampling Design:

A researcher should take into consideration the following aspects

while developing a sample design:

32
1) Type Of Universe:

The first step involved in developing sample design is to clearly
define the number of cases, technically known as the universe. A universe
may be finite or infinite. In a finite universe the number of items is certain,
whereas in the case of an infinite universe the number of items is infinite
(i.e., there is no idea about the total number of items). For example, while
the population of a city or the number of workers in a factory comprise
finite universes, the number of stars in the sky, or throwing of a dice
represent infinite universe.

2) Sampling Unit:

Prior to selecting a sample, decision has to be made about the
sampling unit. A sampling unit may be a geographical area like a state,
district, village, etc., or a social unit like a family, religious community,
school, etc., or it may also be an individual. At times, the researcher would
have to choose one or more of such units for his/her study.

3) Source List:

Source list is also known as the ‘sampling frame’, from which the
sample is to be selected. The source list consists of names of all the items
of a universe. The researcher has to prepare a source list when it is not
available. The source list must be reliable, comprehensive, correct, and
appropriate. It is important that the source list should be as representative
of the population as possible.

4) Size Of Sample:

Size of the sample refers to the number of items to be chosen from
the universe to form a sample. For a researcher, this constitutes a major
problem. The size of sample must be optimum. An optimum sample may
be defined as the one that satisfies the requirements of representativeness,
flexibility, efficiency, and reliability. While deciding the size of sample,
a researcher should determine the desired precision and the acceptable
confidence level for the estimate. The size of the population variance
should be considered, because in the case of a larger variance generally a
larger sample is required. The size of the population should be considered,
33
as it also limits the sample size. The parameters of interest in a research
study should also be considered, while deciding the sample size. Besides,
costs or budgetary constraint also plays a crucial role in deciding the
sample size.

(A) Parameters Of Interest:

The specific population parameters of interest should also be

considered while determining the sample design. For example, the
researcher may want to make an estimate of the proportion of persons with
certain characteristic in the population, or may be interested in knowing
some average regarding the population. The population may also consist
of important sub-groups about whom the researcher would like to make
estimates. All such factors have strong impact on the sample design the
researcher selects.

(B) Budgetary Constraint:

From the practical point of view, cost considerations exercise a

major influence on the decisions related to not only the sample size, but
also on the type of sample selected. Thus, budgetary constraint could also
lead to the adoption of a non-probability sample design.

(c) Sampling Procedure:

Finally, the researcher should decide the type of sample or the

technique to be adopted for selecting the items for a sample. This technique
or procedure itself may represent the sample design. There are different
sample designs from which a researcher should select one for his/her
study. It is clear that the researcher should select that design which, for a
given sample size and budget constraint, involves a smaller error.

7.2 Criteria For Selecting A Sampling Procedure:

Basically, two costs are involved in a sampling analysis, which

govern the selection of a sampling procedure. They are:

1) The cost of data collection, and

2) The cost of drawing incorrect inference from the selected data.
34
 There are two causes of incorrect inferences, namely systematic
bias and sampling error. Systematic bias arises out of errors in the
sampling procedure. They cannot be reduced or eliminated by increasing
the sample size. Utmost, the causes of these errors can be identified and
corrected. Generally, a systematic bias arises out of one or more of the
following factors:

a. Inappropriate sampling frame,

b. Defective measuring device,
c. Non-respondents,
d. Indeterminacy principle, and
e. Natural bias in the reporting of data.

Sampling error refers to the random variations in the sample

estimates around the true population parameters. Because they occur
randomly and likely to be equally in either direction, they are of
compensatory type, the expected value of which errors tend to be equal
to zero. Sampling error tends to decrease with the increase in the size of
the sample. It also becomes smaller in magnitude when the population is
homogenous.

Sampling error can be computed for a given sample size and

design. The measurement of sampling error is known as ‘precision of the
sampling plan’. When the sample size is increased, the precision can be
improved. However, increasing the sample size has its own limitations.
The large sized sample not only increases the cost of data collection, but
also increases the systematic bias. Thus, an effective way of increasing
the precision is generally to choose a better sampling design, which has
smaller sampling error for a given sample size at a specified cost. In
practice, however, researchers generally prefer a less precise design owing
to the ease in adopting the same, in addition to the fact that systematic bias
can be controlled better way in such designs.

In sum, while selecting the sample, a researcher should ensure that

the procedure adopted involves a relatively smaller sampling error and
helps to control systematic bias.

35
7.3 Characteristics Of A Good Sample Design:

The following are the characteristic features of a good sample

design:

a. The sample design should yield a truly representative sample;

b. The sample design should be such that it results in small sampling
error;
c. The sample design should be viable in the context of budgetary
constraints of the research study;
d. The sample design should be such that the systematic bias can be
controlled; and
e. The sample must be such that the results of the sample study would
be applicable, in general, to the universe at a reasonable level of
confidence.

7.4 Different Types Of Sample Designs:

Sample designs may be classified into different categories based

on two factors, namely, the representation basis and the element selection
technique. Under the representation basis, the sample may be classified as:

I. Non-probability sampling
II. Probability sampling

While probability sampling is based on random selection, the non-

probability sampling is based on ‘non-random’ selection of samples.

I. Non-Probability Sampling:

Non-probability sampling is the sampling procedure that does

not afford any basis for estimating the probability that each item in the
population would have an equal chance of being included in the sample.
Non-probability sampling is also known as deliberate sampling, judgment
sampling and purposive sampling. Under this type of sampling, the items
for the sample are deliberately chosen by the researcher; and his/her choice
concerning the choice of items remains supreme. In other words, under
non-probability sampling the researchers select a particular unit of the
universe for forming a sample on the basis that the small number that
36
is thus selected out of a huge one would be typical or representative of
the whole population. For example, to study the economic conditions of
people living in a state, a few towns or village may be purposively selected
for an intensive study based on the principle that they are representative of
the entire state. In such a case, the judgment of the researcher of the study
assumes prime importance in this sampling design.

Quota Sampling:

Quota sampling is also an example of non-probability sampling.

Under this sampling, the researchers simply assume quotas to be filled
from different strata, with certain restrictions imposed on how they should
be selected. This type of sampling is very convenient and is relatively less
expensive. However, the samples selected using this method certainly do
not satisfy the characteristics of random samples. They are essentially
judgment samples and inferences drawn based on that, would not be
amenable to statistical treatment in a formal way.

II. Probability Sampling:

Probability sampling is also known as ‘choice sampling’ or ‘random

sampling’. Under this sampling design, every item of the universe has
an equal chance of being included in the sample. In a way, it is a lottery
method under which individual units are selected from the whole group,
not deliberately, but by using some mechanical process. Therefore, only
chance would determine whether an item or the other would be included
in the sample or not. The results obtained from probability or random
sampling would be assured in terms of probability. That is, the researcher
can measure the errors of estimation or the significance of results obtained
from the random sample. This is the superiority of random sampling design
over the deliberate sampling design. Random sampling satisfies the law of
statistical regularity, according to which if on an average the sample chosen
is random, then it would have the same composition and characteristics
of the universe. This is the reason why the random sampling method is
considered the best technique of choosing a representative sample.

37
The following are the implications of the random sampling:
i. it provides each element in the population an equal probable chance
of being chosen in the sample, with all choices being independent of
one another and
ii. it offers each possible sample combination an equal probable
opportunity of being selected.

7.5 Method Of Selecting A Random Sample:

The process of selecting a random sample involves writing the

name of each element of a finite population on a slip of paper and putting
them into a box or a bag. Then they have to be thoroughly mixed and
then the required number of slips for the sample can be picked one after
the other without replacement. While doing this, it has to be ensured that
in successive drawings each of the remaining elements of the population
has an equal chance of being chosen. This method results in the same
probability for each possible sample.

7.6 Complex Random Sampling Designs:

Under restricted sampling technique, the probability sampling

may result in complex random sampling designs. Such designs are
known as mixed sampling designs. Many of such designs may represent
a combination of non-probability and probability sampling procedures in
choosing a sample.

Some of the prominent complex random sampling designs are as follows:

(i) Systematic Sampling:

In some cases, the best way of sampling is to select every first

item on a list. Sampling of this kind is called as systematic sampling. An
element of randomness is introduced in this type of sampling by using
random numbers to select the unit with which to start. For example, if
a 10 per cent sample is required out of 100 items, the first item would be
selected randomly from the first low of item and thereafter every 10th
item. In this kind of sampling, only the first unit is selected randomly,
while rest of the units of the sample is chosen at fixed intervals.

38
(ii) Stratified Sampling:

When a population from which a sample is to be selected does
not comprise a homogeneous group, stratified sampling technique is
generally employed for obtaining a representative sample. Under stratified
sampling, the population is divided into many sub-populations in such a
manner that they are individually more homogeneous than the rest of the
total population. Then, items are selected from each stratum to form a
sample. As each stratum is more homogeneous than the remaining total
population, the researcher is able to obtain a more precise estimate for each
stratum and by estimating each of the component parts more accurately;
he/she is able to obtain a better estimate of the whole. In sum, stratified
sampling method yields more reliable and detailed information.

(iii) Cluster Sampling:

When the total area of research interest is large, a convenient way

in which a sample can be selected is to divide the area into a number of
smaller non-overlapping areas and then randomly selecting a number of
such smaller areas. In the process, the ultimate sample would consist of all
the units in these small areas or clusters. Thus in cluster sampling, the total
population is sub-divided into numerous relatively smaller subdivisions,
which in themselves constitute clusters of still smaller units. And then,
some of such clusters are randomly chosen for inclusion in the overall
sample.

(iv) Area Sampling:

When clusters are in the form of some geographic subdivisions,

then cluster sampling is termed as area sampling. That is, when the
primary sampling unit represents a cluster of units based on geographic
area, the cluster designs are distinguished as area sampling. The merits
and demerits of cluster sampling are equally applicable to area sampling.

(v) Multi-Stage Sampling:

A further development of the principle of cluster sampling is

multi-stage sampling. When the researcher desires to investigate the
working efficiency of nationalized banks in India and a sample of few
39
banks is required for this purpose, the first stage would be to select large
primary sampling unit like the states in the country. Next, certain districts
may be selected and all banks interviewed in the chosen districts. This
represents a two-stage sampling design, with the ultimate sampling units
being clusters of districts.

On the other hand, if instead of taking census of all banks within

the selected districts, the researcher chooses certain towns and interviews
all banks in it, this would represent three-stage sampling design. Again,
if instead of taking a census of all banks within the selected towns, the
researcher randomly selects sample banks from each selected town, then
it represents a case of using a four-stage sampling plan. Thus, if the
researcher selects randomly at all stages, then it is called as multi-stage
random sampling design.

(vi) Sampling With Probability Proportional To Size:

When the case of cluster sampling units does not have exactly or
approximately the same number of elements, it is better for the researcher
to adopt a random selection process, where the probability of inclusion
of each cluster in the sample tends to be proportional to the size of the
cluster. For this, the number of elements in each cluster has to be listed,
irrespective of the method used for ordering it. Then the researcher should
systematically pick the required number of elements from the cumulative
totals. The actual numbers thus chosen would not however reflect the
individual elements, but would indicate as to which cluster and how many
from them are to be chosen by using simple random sampling or systematic
sampling. The outcome of such sampling is equivalent to that of simple
random sample. The method is also less cumbersome and is also relatively
less expensive.

Thus, a researcher has to pass through various stages of conducting

research once the problem of interest has been selected. Research
methodology familiarizes a researcher with the complex scientific methods
of conducting research, which yield reliable results that are useful to
policy-makers, government, industries etc. in decision-making.

40
End of Chapter Questions:[IMPORTANT QUESTIONS FOR EXAM]

1. Define research.
2. What are the objectives of research?
3. State the significance of research.
4. What is the importance of knowing how to do research?
5. Briefly outline research process.
6. Highlight the different research approaches.
7. Discuss the qualities of a researcher.
8. Explain the different types of research.
9. What is a research problem?
10. Outline the features of research design.
11. Discuss the features of a good research design.
12. Describe the different types of research design.
13. Explain the significance of research design.
14. What is a case study?
15. Discuss the criteria for evaluating case study.
16. Define hypothesis.
17. What are the characteristic features of a hypothesis?
18. Distinguish between null and alternative hypothesis.
19. Differentiate type i error and type ii error.
20. How is a hypothesis tested?
21. Define the concept of sampling design.

41
22. Describe the steps involved in sampling design.
23. Discuss the criteria for selecting a sampling procedure.
24. Distinguish between probability and non-probability sampling.
25. How is a random sample selected?
26. Explain complex random sampling designs.

***

42
 CHAPTER—II

Lesson 1: Data Collection & Sources Of Data

1. Primary Data, Secondary Data

2. Investigation
3. Indirect Oral Methods Of Collecting Primary Data
4. Direct Personal Interviews
5. Information Received Through Local Agencies
6. Mailed Questionnaire Method
7. Schedules Sent Through Enumerators

Meaning of primary data, Secondary data,Preliminaries of data

collection,Method of data collection,Methods of collecting primary
data ,Usefulness of primary data ,Merits and demerits of
different methods of primary data collection
Precautions while collecting primary data.
Secondary data collection

Introduction:

It is important for a researcher to know the sources of data which

he requires for different purposes. Data are nothing but the information.
There are two sources of information or data they are - Primary and
Secondary data. The data are name after the source. Primary data refers
to the data collected for the first time, whereas secondary data refers to
the data that have already been collected and used earlier by somebody
or some agency. For example, the statistics collected by the Government

43
of India relating to the population is primary data for the Government of
India since it has been collected for the first time. Later when the same
data are used by a researcher for his study of a particular problem, then
the same data become the secondary data for the researcher. Both the
sources of information have their merits and demerits. The selection of
a particular source depends upon the (a) purpose and scope of enquiry,
(b) availability of time, (c) availability of finance, (d) accuracy required,
(e) statistical tools to be used, (f) sources of information (data), and (g)
method of data collection.

(a) Purpose And Scope Of Enquiry:

The purpose and scope of data collection or survey should be

clearly set out at the very beginning. It requires the clear statement of the
problem indicating the type of information which is needed and the use for
which it is needed. If for example, the researcher is interested in knowing
the nature of price change over a period of time, it would be necessary to
collect data of commodity prices. It must be decided whether it would be
helpful to study wholesale or retail prices and the possible uses to which
such information could be put. The objective of an enquiry may be either
to collect specific information relating to a problem or adequate data to
test a hypothesis. Failure to set out clearly the purpose of enquiry is bound
to lead to confusion and waste of resources.

After the purpose of enquiry has been clearly defined, the next step
is to decide about the scope of the enquiry. Scope of the enquiry means the
coverage with regard to the type of information, the subject-matter and
geographical area. For instance, an enquiry may relate to India as a whole
or a state or an industrial town wherein a particular problem related to a
particular industry can be studied.

(b) Availability Of Time:

The investigation should be carried out within a reasonable period

of time, failing which the information collected may become outdated,
and would have no meaning at all. For instance, if a producer wants to
know the expected demand for a product newly launched by him and the
result of the enquiry that the demand would be meager takes two years
to reach him, then the whole purpose of enquiry would become useless
44
because by that time he would have already incurred a huge loss. Thus, in
this respect the information is quickly required and hence the researcher
has to choose the type of enquiry accordingly.

(c) Availability Of Resources:

The investigation will greatly depend on the resources available like

number of skilled personnel, the financial position etc. If the number of
skilled personnel who will carry out the enquiry is quite sufficient and the
availability of funds is not a problem, then enquiry can be conducted over
a big area covering a good number of samples, otherwise a small sample
size will do.

(d) The Degree Of Accuracy Desired:

Deciding the degree of accuracy required is a must for the

investigator, because absolute accuracy in statistical work is seldom
achieved. This is so because (i) statistics are based on estimates, (ii) tools
of measurement are not always perfect and (iii) there may be unintentional
bias on the part of the investigator, enumerator or informant. Therefore, a
desire of 100% accuracy is bound to remain unfulfilled. Degree of accuracy
desired primarily depends upon the object of enquiry. For example, when
we buy gold, even a difference of 1/10th gram in its weight is significant,
whereas the same will not be the case when we buy rice or wheat. However,
the researcher must aim at attaining a higher degree of accuracy, otherwise
the whole purpose of research would become meaningless.

(e) Statistical Tools To Be Used:

A well defined and identifiable object or a group of objects with

which the measurements or counts in any statistical investigation are
associated is called a statistical unit. For example, in socio-economic survey
the unit may be an individual, a family, a household or a block of locality.
A very important step before the collection of data begins is to define
clearly the statistical units on which the data are to be collected. In number
of situations the units are conventionally fixed like the physical units of
measurement, such as meters, kilometers, quintals, hours, days, weeks etc.,
which are well defined and do not need any elaboration or explanation.
However, in many statistical investigations, particularly relating to socio-
45
economic studies, arbitrary units are used which must be clearly defined.
This is a must because in the absence of a clear cut and precise definition of
the statistical units, serious errors in the data collection may be committed
in the sense that we may collect irrelevant data on the items, which should
have, in fact, been excluded and omit data on certain items which should
have been included. This will ultimately lead to fallacious conclusions.

(f) Sources Of Information (Data):

After deciding about the unit, a researcher has to decide about the
source from which the information can be obtained or collected. For any
statistical inquiry, the investigator may collect the data first hand or he
may use the data from other published sources, such as publications of the
government/semi-government organizations or journals and magazines
etc.

(g) Method Of Data Collection:

There is no problem if secondary data are used for research.

However, if primary data are to be collected, a decision has to be taken
whether (i) census method or (ii) sampling technique is to be used for
data collection. In census method, we go for total enumeration i.e., all
the units of a universe have to be investigated. But in sampling technique,
we inspect or study only a selected representative and adequate fraction
of the population and after analyzing the results of the sample data we
draw conclusions about the characteristics of the population. Selection
of a particular technique becomes difficult because where population
or census method is more scientific and 100% accuracy can be attained
through this method, choosing this becomes difficult because it is time
taking, it requires more labor and it is very expensive. Therefore, for a
single researcher or for a small institution it proves to be unsuitable. On
the other hand, sample method is less time taking, less laborious and less
expensive but a 100% accuracy cannot be attained through this method
because of sampling and non-sampling errors attached to this method.
Hence, a researcher has to be very cautious and careful while choosing a
particular method.

46
Methods of Collecting Primary Data:

Primary data may be obtained by applying any of the following

methods:

1. Direct Personal Interviews.

2. Indirect Oral Interviews.
3. Information from Correspondents.
4. Mailed Questionnaire Methods.
5. Schedule Sent Through Enumerators.

1. Direct Personal Interviews:

A face to face contact is made with the informants (persons from

whom the information is to be obtained) under this method of collecting
data. The interviewer asks them questions pertaining to the survey and
collects the desired information. Thus, if a person wants to collect data
about the working conditions of the workers of the Tata Iron and Steel
Company, Jamshedpur, he would go to the factory, contact the workers
and obtain the desired information. The information collected in this
manner is first hand and also original in character. There are many merits
and demerits of this method, which are discussed as under:

Merits:

1. Most often respondents are happy to pass on the information required

from them when contacted personally and thus response is encouraging.
2. The information collected through this method is normally more
accurate because interviewer can clear doubts of the informants
about certain questions and thus obtain correct information. In case
the interviewer apprehends that the informant is not giving accurate
information, he may cross-examine him and thereby try to obtain the
information.
3. This method also provides the scope for getting supplementary
information from the informant, because while interviewing it is
possible to ask some supplementary questions which may be of greater
use later.
4. There might be some questions which the interviewer would find
difficult to ask directly, but with some tactfulness, he can mingle such
47
 questions with others and get the desired information. He can twist
the questions keeping in mind the informant’s reaction. Precisely, a
delicate situation can usually he handled more effectively by a personal
interview than by other survey techniques.
5. The interviewer can adjust the language according to the status and
educational level of the person interviewed, and thereby can avoid
inconvenience and misinterpretation on the part of the informant.

Demerits:

1. This method can prove to be expensive if the number of informants is

large and the area is widely spread.
2. There is a greater chance of personal bias and prejudice under this
method as compared to other methods.
3. The interviewers have to be thoroughly trained and experienced;
otherwise they may not be able to obtain the desired information.
Untrained or poorly trained interviewers may spoil the entire work.
4. This method is more time taking as compared to others. This is because
interviews can be held only at the convenience of the informants.
Thus, if information is to be obtained from the working members
of households, interviews will have to be held in the evening or on
week end. Even during evening only an hour or two can be used for
interviews and hence, the work may have to be continued for a long
time, or a large number of people may have to be employed which may
involve huge expenses.

Conclusion:

Though there are some demerits in this method of data collection

still we cannot say that it is not useful. The matter of fact is that this
method is suitable for intensive rather than extensive field surveys. Hence,
it should be used only in those cases where intensive study of a limited
field is desired.

In the present time of extreme advancement in the communication

system, the investigator instead of going personally and conducting a face
to face interview may also obtain information over telephone. A good
number of surveys are being conducted every day by newspapers and
television channels by sending the reply either by e-mail or SMS. This
48
method has become very popular nowadays as it is less expensive and the
response is extremely quick. But this method suffers from some serious
defects, such as (a) those who own a phone or a television only can be
approached by this method, (b) only few questions can be asked over
phone or through television, (c) the respondents may give a vague and
reckless answers because answers on phone or through SMS would have
to be very short.

2. Indirect Oral Interviews:

Under this method of data collection, the investigator contacts third

parties generally called ‘witnesses’ who are capable of supplying necessary
information. This method is generally adopted when the information
to be obtained is of a complex nature and informants are not inclined
to respond if approached directly. For example, when the researcher is
trying to obtain data on drug addiction or the habit of taking liquor, there
is high probability that the addicted person will not provide the desired
data and hence will disturb the whole research process. In this situation
taking the help of such persons or agencies or the neighbours who know
them well becomes necessary. Since these people know the person well,
they can provide the desired data. Enquiry Committees and Commissions
appointed by the Government generally adopt this method to get people’s
views and all possible details of the facts related to the enquiry.

Though this method is very popular, its correctness depends upon a

number of factors such as

1. The person or persons or agency whose help is solicited must be of

proven integrity; otherwise any bias or prejudice on their part will not
bring out the correct information and the whole process of research
will become useless.
2. The ability of the interviewers to draw information from witnesses by
means of appropriate questions and cross-examination.
3. It might happen that because of bribery, nepotism or certain other
reasons those who are collecting the information give it such a twist
that correct conclusions are not arrived at.

Therefore, for the success of this method it is necessary that the

evidence of one person alone is not relied upon. Views from other persons
49
and related agencies should also be ascertained to find the real position
.Utmost care must be exercised in the selection of these persons because it
is on their views that the final conclusions are reached.

3. Information from Correspondents:

The investigator appoints local agents or correspondents in different

places to collect information under this method. These correspondents
collect and transmit the information to the central office where data are
processed. This method is generally adopted by news paper agencies.
Correspondents who are posted at different places supply information
relating to such events as accidents, riots, strikes, etc., to the head office.
The correspondents are generally paid staff or sometimes they may be
honorary correspondents also. This method is also adopted generally by
the government departments in such cases where regular information
is to be collected from a wide area. For example, in the construction of
a wholesale price index numbers regular information is obtained from
correspondents appointed in different areas. The biggest advantage of this
method is that, it is cheap and appropriate for extensive investigation. But
a word of caution is that it may not always ensure accurate results because
of the personal prejudice and bias of the correspondents. As stated earlier,
this method is suitable and adopted in those cases where the information
is to be obtained at regular intervals from a wide area.

4. Mailed Questionnaire Method:

Under this method, a list of questions pertaining to the survey

which is known as ‘Questionnaire’ is prepared and sent to the various
informants by post. Sometimes the researcher himself too contacts the
respondents and gets the responses related to various questions in the
questionnaire. The questionnaire contains questions and provides space
for answers. A request is made to the informants through a covering letter
to fill up the questionnaire and send it back within a specified time. The
questionnaire studies can be classified on the basis of:

i. The degree to which the questionnaire is formalized or structured.

ii. The disguise or lack of disguise of the questionnaire and
iii. The communication method used.

50
 When no formal questionnaire is used, interviewers adapt their
questioning to each interview as it progresses. They might even try to elicit
responses by indirect methods, such as showing pictures on which the
respondent comments. When a researcher follows a prescribed sequence
of questions, it is referred to as structured study. On the other hand, when
no prescribed sequence of questions exists, the study is non-structured.

When questionnaires are constructed in such a way that the

objective is clear to the respondents then these questionnaires are known
as non- disguised; on the other hand, when the objective is not clear, the
questionnaire is a disguised one. On the basis of these two classifications,
four types of studies can be distinguished:

1. Non-disguised structured,
2. Non-disguised non-structured,
3. Disguised structured and
4. Disguised non-structured.

There are certain merits and demerits of this method of data collection
which are discussed below:

Merits:

1. Questionnaire method of data collection can be easily adopted where

the field of investigation is very vast and the informants are spread
over a wide geographical area.
2. This method is relatively cheap and expeditious provided the informants
respond in time.
3. This method has proved to be superior when compared to other
methods like personal interviews or telephone method. This is because
when questions pertaining to personal nature or the ones requiring
reaction by the family are put forth to the informants, there is a chance
for them to be embarrassed in answering them.

Demerits:

1. This method can be adopted only where the informants are literates
so that they can understand written questions and lend the answers in
writing.
51
2. It involves some uncertainty about the response. Co-operation on the
part of informants may be difficult to presume.
3. The information provided by the informants may not be correct and it
may be difficult to verify the accuracy.

However, by following the guidelines given below, this method can be

made more effective:

The questionnaires should be made in such a manner that they

do not become an undue burden on the respondents; otherwise the
respondents may not return them back.

i. Prepaid postage stamp should be affixed

ii. The sample should be large
iii. It should be adopted in such enquiries where it is expected that
the respondents would return the questionnaire because of their
own interest in the enquiry.
iv. It should be preferred in such enquiries where there could be a
legal compulsion to provide the information.

5. Schedules Sent Through Enumerators:

Another method of data collection is sending schedules through

the enumerators or interviewers. The enumerators contact the informants,
get replies to the questions contained in a schedule and fill them in their
own handwriting in the questionnaire form. There is difference between
questionnaire and schedule. Questionnaire refers to a device for securing
answers to questions by using a form which the respondent fills in him self,
whereas schedule is the name usually applied to a set of questions which
are asked in a face-to face situation with another person. This method is
free from most of the limitations of the mailed questionnaire method.

Merits:

The main merits or advantages of this method are listed below:

1. It can be adopted in those cases where informants are illiterate.

2. There is very little scope of non-response as the enumerators go
personally to obtain the information.
52
3. The information received is more reliable as the accuracy of statements
can be checked by supplementary questions wherever necessary.

This method too like others is not free from defects or limitations. The
main limitations are listed below:

Demerits:

1. In comparison to other methods of collecting primary data, this

method is quite costly as enumerators are generally paid persons.
2. The success of the method depends largely upon the training imparted
to the enumerators.
3. Interviewing is a very skilled work and it requires experience
and training. Many statisticians have the tendency to neglect this
extremely important part of the data collecting process and this result
in bad interviews. Without good interviewing most of the information
collected may be of doubtful value.
4. Interviewing is not only a skilled work but it also requires a great
degree of politeness and thus the way the enumerators conduct the
interview would affect the data collected. When questions are asked
by a number of different interviewers, it is possible that variations in
the personalities of the interviewers will cause variation in the answers
obtained. This variation will not be obvious. Hence, every effort must
be made to remove as much of variation as possible due to different
interviewers.

Secondary Data:

As stated earlier, secondary data are those data which have

already been collected and analyzed by some earlier agency for its own
use, and later the same data are used by a different agency. According to
W.A.Neiswanger, “A primary source is a publication in which the data are
published by the same authority which gathered and analyzed them. A
secondary source is a publication, reporting the data which was gathered
by other authorities and for which others are responsible.”

53
Sources Of Secondary Data:

The various sources of secondary data can be divided into two

broad categories:
1. Published sources, and
2. Unpublished sources.

1. Published Sources:

The governmental, international and local agencies publish

statistical data, and chief among them are explained below:

(a) International Bublications:

There are some international institutions and bodies like I.M.F,
I.B.R.D, I.C.A.F.E and U.N.O who publish regular and occasional reports
on economic and statistical matters.

(b) Official Publications of Central and State Governments:

Several departments of the Central and State Governments

regularly publish reports on a number of subjects. They gather additional
information. Some of the important publications are: The Reserve Bank of
India Bulletin, Census of India, Statistical Abstracts of States, Agricultural
Statistics of India, Indian Trade Journal, etc.

(c) Semi-Official Publications:

Semi-Government institutions like Municipal Corporations,

District Boards, Panchayats, etc. Publish reports relating to different
matters of public concern.

(d) Publications of Research Institutions:

Indian Statistical Institute (I.S.I), Indian Council of Agricultural

Research (I.C.A.R), Indian Agricultural Statistics Research Institute
(I.A.S.R.I), etc. Publish the findings of their research programmes.

(e) Publications of various Commercial and Financial Institutions

54
(f) Reports of various Committees and Commissions appointed by the
Government as the Raj Committee’s Report on Agricultural Taxation,
Wanchoo Committee’s Report on Taxation and Black Money, etc. Are also
important sources of secondary data.

(g) Journals and News Papers:

Journals and News Papers are very important and powerful source
of secondary data. Current and important materials on statistics and socio-
economic problems can be obtained from journals and newspapers like
Economic Times, Commerce, Capital, Indian Finance, Monthly Statistics
of trade etc.

2. Unpublished Sources:

Unpublished data can be obtained from many unpublished sources

like records maintained by various government and private offices, the
theses of the numerous research scholars in the universities or institutions
etc.

Precautions In The Use Of Secondary Data:

Since secondary data have already been obtained, it is highly

desirable that a proper scrutiny of such data is made before they are
used by the investigator. In fact the user has to be extra-cautious while
using secondary data. In this context Prof. Bowley rightly points out that
“Secondary data should not be accepted at their face value.” The reason
being that data may be erroneous in many respects due to bias, inadequate
size of the sample, substitution, errors of definition, arithmetical errors
etc. Even if there is no error such data may not be suitable and adequate
for the purpose of the enquiry. Prof. SimonKuznet’s view in this regard is
also of great importance. According to him, “the degree of reliability of
secondary source is to be assessed from the source, the compiler and his
capacity to produce correct statistics and the users also, for the most part,
tend to accept a series particularly one issued by a government agency at
its face value without enquiring its reliability”.

Therefore, before using the secondary data the investigators should

consider the following factors:
55
4. The Suitability Of Data:

The investigator must satisfy himself that the data available are
suitable for the purpose of enquiry. It can be judged by the nature and
scope of the present enquiry with the original enquiry. For example, if the
object of the present enquiry is to study the trend in retail prices, and if the
data provide only wholesale prices, such data are unsuitable.

(A) Adequacy Of Data:

If the data are suitable for the purpose of investigation then we must
consider whether the data are useful or adequate for the present analysis.
It can be studied by the geographical area covered by the original enquiry.
The time for which data are available is very important element. In the
above example, if our object is to study the retail price trend of india, and
if the available data cover only the retail price trend in the state of bihar,
then it would not serve the purpose.

(b) Reliability Of Data:

The reliability of data is must. Without which there is no meaning in

research. The reliability of data can be tested by finding out the agency that
collected such data. If the agency has used proper methods in collection of
data, statistics may be relied upon.

It is not enough to have baskets of data in hand. In fact, data in

a raw form are nothing but a handful of raw material waiting for proper
processing so that they can become useful. Once data have been obtained
from primary or secondary source, the next step in a statistical investigation
is to edit the data i.e. To scrutinize the same. The chief objective of editing
is to detect possible errors and irregularities. The task of editing is a
highly specialized one and requires great care and attention. Negligence
in this respect may render useless the findings of an otherwise valuable
study. Editing data collected from internal records and published sources
is relatively simple but the data collected from a survey need excessive
editing.
While editing primary data, the following considerations should be borne
in mind:
1. The data should be complete in every respect
56
 2. The data should be accurate
3. The data should be consistent, and
4. The data should be homogeneous.

Data to posses the above mentioned characteristics have to undergo the

same type of editing which is discussed below:

5. Editing for Completeness:

while editing, the editor should see that each schedule and
questionnaire is complete in all respects. He should see to it that the answers
to each and every question have been furnished. If some questions are not
answered and if they are of vital importance, the informants should be
contacted again either personally or through correspondence. Even after
all the efforts it may happen that a few questions remain unanswered. In
such questions, the editor should mark ‘No answer’ in the space provided
for answers and if the questions are of vital importance then the schedule
or questionnaire should be dropped.

(a) Editing for Consistency:

At the time of editing the data for consistency, the editor should see
that the answers to questions are not contradictory in nature. If they are
mutually contradictory answers, he should try to obtain the correct answers
either by referring back the questionnaire or by contacting, wherever
possible, the informant in person. For example, if amongst others, two
questions in questionnaire are (a) Are you a student? (b) Which class do
you study and the reply to the first question is ‘no’ and to the latter ‘tenth’
then there is contradiction and it should be clarified.

(b) Editing for Accuracy:

The reliability of conclusions depends basically on the correctness

of information. If the information supplied is wrong, conclusions can never
be valid. It is, therefore, necessary for the editor to see that the information
is accurate in all respects. If the inaccuracy is due to arithmetical errors, it
can be easily detected and corrected. But if the cause of inaccuracy is faulty
information supplied, it may be difficult to verify it and an example of this
kind is information relating to income, age etc.
57
(c) Editing For Homogeneity:

Homogeneity means the condition in which all the questions have

been understood in the same sense. The editor must check all the questions
for uniform interpretation. For example, as to the question of income, if
some informants have given monthly income, others annual income and
still others weekly income or even daily income, no comparison can be
made. Therefore, it becomes an essential duty of the editor to check up
that the information supplied by the various people is homogeneous and
uniform.

Choice Between Primary and Secondary Data:

As we have already seen, there are a lot of differences in the methods

of collecting Primary and Secondary data. Primary data which is to be
collected originally involves an entire scheme of plan starting with the
definitions of various terms used, units to be employed, type of enquiry
to be conducted, extent of accuracy aimed at etc. For the collection of
secondary data, a mere compilation of the existing data would be sufficient.
A proper choice between the type of data needed for any particular
statistical investigation is to be made after taking into consideration the
nature, objective and scope of the enquiry; the time and the finances at the
disposal of the agency; the degree of precision aimed at and the status of
the agency (whether government- state or central-or private institution of
an individual).

In using the secondary data, it is best to obtain the data from the
primary source as far as possible. By doing so, we would at least save
ourselves from the errors of transcription which might have inadvertently
crept in the secondary source. Moreover, the primary source will also
provide us with detailed discussion about the terminology used, statistical
units employed, size of the sample and the technique of sampling (if
sampling method was used), methods of data collection and analysis of
results and we can ascertain ourselves if these would suit our purpose.
Now-a-days in a large number of statistical enquiries, secondary data
are generally used because fairly reliable published data on a large number
of diverse fields are now available in the publications of governments,
private organizations and research institutions, agencies, periodicals and
magazines etc. In fact, primary data are collected only if there do not exist
58
any secondary data suited to the investigation under study. In some of the
investigations both primary as well as secondary data may be used.

Summary:

There are two types of data, primary and secondary. Data which are
collected first hand are called Primary data and data which have already
been collected and used by somebody are called Secondary data. There are
two methods of collecting data: (a) Survey method or total enumeration
method and (b) Sample method. When a researcher goes for investigating
all the units of the subject, it is called as survey method. On the other hand
if he/she resorts to investigating only a few units of the subject and gives
the result on the basis of that, it is known as sample survey method. There
are different sources of collecting Primary and Secondary data. Some of
the important sources of Primary data are—Direct Personal Interviews,
Indirect Oral Interviews, Information from Correspondents, Mailed
questionnaire method, Schedules sent through enumerators and so on.
Though all these sources or methods of Primary data have their relative
merits and demerits, a researcher should use a particular method with lot
of care. There are basically two sources of collecting secondary data- (a)
Published sources and (b) Unpublished sources. Published sources are like
publications of different government and semi-government departments,
research institutions and agencies etc. Whereas unpublished sources
are like records maintained by different government departments and
unpublished theses of different universities etc. Editing of secondary data
is necessary for different purposes as – editing for completeness, editing
for consistency, editing for accuracy and editing for homogeneity.

It is always a tough task for the researcher to choose between

primary and secondary data. Though primary data are more authentic
and accurate, time, money and labor involved in obtaining these more
often prompt the researcher to go for the secondary data. There are certain
amount of doubt about its authenticity and suitability, but after the arrival
of many government and semi government agencies and some private
institutions in the field of data collection, most of the apprehensions in
the mind of the researcher have been removed.

59
Self Assesment Questions (SAQS):

a) Explain primary and secondary data and distinguish between them.

(Refer the introduction part of this lesson.)
b) Explain the different methods of collecting primary data.
(Explain direct personal, indirect oral interview, information received
through agencies etc.)
c) Explain the merits and demerits of different methods of
collecting primary data.
(Refer the methods of collecting primary data)
d) Explain the different sources of secondary data and the
precautions in using secondary data.
e) What is editing of secondary data? Why is it required?
f) What are the different types of editing of secondary data?

GLOSSARY OF TERMS:

Primary Source:
It is one that itself collects the data.

Secondary Source:
It is one that makes available data collected by some other agency.

Collection of Statistics:
Collection means the assembling for the purpose of particular
investigation of entirely new data presumably not already available in
published sources.

Questionnaire:
A list of questions properly selected and arranged pertaining to the
investigation.

Investigator:
Investigator is a person who collects the information.

Respondent:
A person who fills the questionnaire or provides the required
information.
***
60
 CHAPTER II

Lesson 2: Questionnaire & Sampling

111 Meaning Of Questionnaire

Drafting Of Questionnaire.
Size Of Questions
Clarity Of Questions
Logical Sequence Of Questions
Simple Meaning Questions
Other Requirements Of A Good Questionnaire
Meaning And Essentials Of Sampling.

Introduction:

Nowadays questionnaire is widely used for data collection in social

research. It is a reasonably fair tool for gathering data from large, diverse,
varied and scattered social groups. The questionnaire is the media of
communication between the investigator and the respondents. According
to Bogardus, a questionnaire is a list of questions sent to a number of
persons for their answers and which obtains standardized results that
can be tabulated and treated statistically. The Dictionary of Statistical
Terms defines it as a “group of or sequence of questions designed to elicit

61
information upon a subject or sequence of subjects from information.” A
questionnaire should be designed or drafted with utmost care and caution
so that all the relevant and essential information for the enquiry may be
collected without any difficulty, ambiguity and vagueness. Drafting of a
good questionnaire is a highly specialized job and requires great care skill,
wisdom, efficiency and experience. No hard and fast rule can be laid down
for designing or framing a questionnaire. However, in this connection, the
following general points may be borne in mind:

1. Size Of The Questionnaire Should Be Small:

A researcher should try his best to keep the number of questions

as small as possible, keeping in view the nature, objectives and scope of
the enquiry. Respondent’s time should not be wasted by asking irrelevant
and unimportant questions. A large number of questions would involve
more work for the investigator and thus result in delay on his part in
collecting and submitting the information. A large number of unnecessary
questions may annoy the respondent and he may refuse to cooperate. A
reasonable questionnaire should contain from 15 to 25 questions at large.
If a still larger number of questions are a must in any enquiry, then the
questionnaire should be divided into various sections or parts.

2. The Questions Should Be Clear:

The questions should be easy, brief, unambiguous, non-offending,

courteous in tone, corroborative in nature and to the point, so that much
scope of guessing is left on the part of the respondents.

3. The Questions Should Be Arranged In A Logical Sequence:

Logical arrangement of questions reduces lot of unnecessary work

on the part of the researcher because it not only facilitates the tabulation
work but also does not leave any chance for omissions or commissions.
For example, to find if a person owns a television, the logical order of
questions would be: Do you own a television? When did you buy it? What
is its make? How much did it cost you? Is its performance satisfactory?
Have you ever got it serviced?

62
4. Questions Should Be Simple To Understand:

The vague words like good, bad, efficient, sufficient, prosperity,
rarely, frequently, reasonable, poor, rich etc., should not be used since
these may be interpreted differently by different persons and as such might
give unreliable and misleading information. Similarly the use of words
having double meaning like price, assets, capital income etc., should also
be avoided.

5. Questions Should Be Comprehensive & Easily Answerable:

Questions should be designed in such a way that they are readily

comprehensible and easy to answer for the respondents. They should not
be tedious nor should they tax the respondents’ memory. At the same time
questions involving mathematical calculations like percentages, ratios etc.,
should not be asked.

6. Questions Of Personal & Sensitive Nature Should Not Be

Asked:

There are some questions which disturb the respondents and he/
she may be shy or irritated by hearing such questions. Therefore, every
effort should be made to avoid such questions. For example, ‘do you
cook yourself or your wife cooks?’ ‘Or do you drink?’ Such questions will
certainly irk the respondents and thus be avoided at any cost. If unavoidable
then highest amount of politeness should be used.

7. Types Of Questions:

Under this head, the questions in the questionnaire may be classified

as follows:

(a) Shut Questions:

Shut questions are those where possible answers are suggested by

the framers of the questionnaire and the respondent is required to tick
one of them. Shut questions can further be subdivided into the following
forms:

63
(i) Simple Alternate Questions:

In this type of questions the respondent has to choose from the two
clear cut alternatives like ‘Yes’ or ‘No’, ‘Right or Wrong’ etc. Such questions
are also called as dichotomous questions. This technique can be applied
with elegance to situations where two clear cut alternatives exist.

(ii) Multiple Choice Questions:

Many a times it becomes difficult to define a clear cut alternative
and accordingly in such a situation additional answers between Yes and
No, like Do not know, No opinion, Occasionally, Casually, Seldom etc.,
are added. For example, in order to find if a person smokes or drinks, the
following multiple choice answers may be used:

Do you smoke?

(a) Yes regularly [ ] (b) No never [ ]

(c) Occasionally [ ] (d) Seldom [ ]

Multiple choice questions are very easy and convenient for

the respondents to answer. Such questions save time and also facilitate
tabulation. This method should be used if only a selected few alternative
answers exist to a particular question.

8. Leading Questions Should Be Avoided:

Questions like ‘why do you use a particular type of car, say Maruti
car’ should preferably be framed into two questions-

(i) which car do you use? (ii) why do you prefer it?

It gives smooth ride [ ]

It gives more mileage [ ]
It is cheaper [ ]
It is maintenance free [ ]

64
9 Cross Checks:

The questionnaire should be so designed as to provide internal

checks on the accuracy of the information supplied by the respondents by
including some connected questions at least with respect to matters which
are fundamental to the enquiry.

10 Pre Testing The Questionnaire:

It would be practical in every sense to try out the questionnaire on

a small scale before using it for the given enquiry on a large scale. This
has been found extremely useful in practice. The given questionnaire can
be improved or modified in the light of the drawbacks, shortcomings and
problems faced by the investigator in the pre test.

11 A Covering Letter:

A covering letter from the organizers of the enquiry should

be enclosed along with the questionnaire for the purposes regarding
definitions, units, concepts used in the questionnaire, for taking the
respondent’s confidence, self addressed envelop in case of mailed
questionnaire, mention about award or incentives for the quick response,
a promise to send a copy of the survey report etc.

SAMPLING

Though sampling is not new, the sampling theory has been

developed recently. People knew or not but they have been using the
sampling technique in their day to day life. For example a house wife tests
a small quantity of rice to see whether it has been well-cooked and gives
the generalized result about the whole rice boiling in the vessel. The result
arrived at is most of the times 100% correct. In another example, when
a doctor wants to examine the blood for any deficiency, takes only a few
drops of blood of the patient and examines. The result arrived at is most
of the times correct and represent the whole amount of blood available in
the body of the patient. In all these cases, by inspecting a few, they simply
believe that the samples give a correct idea about the population. Most of
our decision are based on the examination of a few items only i.e. Sample
studies. In the words of Croxton and Cowdon, “It may be too expensive
65
or too time consuming to attempt either a complete or a nearly complete
coverage in a statistical study. Further to arrive at valid conclusions, it
may not be necessary to enumerate all or nearly all of a population. We
may study a sample drawn from the large population and if that sample is
adequately representative of the population, we should be able to arrive at
valid conclusions.”

According to Rosander, “The sample has many advantages over a

census or complete enumeration. If carefully designed, the sample is not
only considerably cheaper but may give results which are just accurate
and sometimes more accurate than those of a census. Hence a carefully
designed sample may actually be better than a poorly planned and executed
census.”

Merits:

1. It saves time:

Sampling method of data collection saves time because fewer items

are collected and processed. When the results are urgently required, this
method is very helpful.

2. It reduces cost:

Since only a few and selected items are studied in sampling, there is
reduction in cost of money and reduction in terms of man hours.

3. More reliable results can be obtained:

Through sampling, more reliable results can be obtained because

(a) there are fewer chances of sampling statistical errors. If there is
sampling error, it is possible to estimate and control the results.(b) Highly
experienced and trained persons can be employed for scientific processing
and analyzing of relatively limited data and they can use their high technical
knowledge and get more accurate and reliable results.

66
4. It provides more detailed information:

As it saves time, money and labor, more detail information can be

collected in a sample survey.

5. Sometimes only sampling method to depend upon:

Some times it so happens that one has to depend upon sampling

method alone because if the population under study is finite, sampling
method is the only method to be used. For example, if someone’s blood has
to be examined, it will become fatal to take all the blood out from the body
and study depending upon the total enumeration method.

6. Administrative convenience:

The organization and administration of sample survey are easy for

the reasons which have been discussed earlier.

7. More scientific:

Since the methods used to collect data are based on scientific theory
and results obtained can be tested, sampling is a more scientific method of
collecting data.

It is not that sampling is free from demerits or shortcomings. There are

certain shortcomings of this method which are discussed below:

1. Illusory conclusion:

If a sample enquiry is not carefully planned and executed, the

conclusions may be inaccurate and misleading.

2. Sample Not Representative:

To make the sample representative is a difficult task. If a

representative sample is taken from the universe, the result is applicable
to the whole population. If the sample is not representative of the universe
the result may be false and misleading.

67
3. Lack Of Experts:

As there are lack of experts to plan and conduct a sample survey, its
execution and analysis, and its results would be
Unsatisfactory and not trustworthy.

4. Sometimes More Difficult Than Census Method:

Sometimes the sampling plan may be complicated and requires

more money, labor and time than a census method.

5. Personal Bias:

There may be personal biases and prejudices with regard to the

choice of technique and drawing of sampling units.

6. Choice Of Sample Size:

If the size of the sample is not appropriate then it may lead to untrue
characteristics of the population.

7. Conditions Of Complete Coverage:

If the information is required for each and every item of the

universe, then a complete enumeration survey is better.

Essentials of sampling:

In order to reach a clear conclusion, the sampling should possess

the following essentials:

1. It must be representative:

The sample selected should possess the similar characteristics of

the original universe from which it has been drawn.

68
2. Homogeneity:

Selected samples from the universe should have similar nature and
should mot have any difference when compared with the universe.

3. Adequate samples:

In order to have a more reliable and representative result, a good

number of items are to be included in the sample.

4. Optimization:

All efforts should be made to get maximum results both in terms

of cost as well as efficiency. If the size of the sample is larger, there is
better efficiency and at the same time the cost is more. A proper size of
sample is maintained in order to have optimized results in terms of cost
and efficiency.

STATISTICAL LAWS

One of the basic reasons for undertaking a sample survey is to

predict and generalize the results for the population as a whole. The logical
process of drawing general conclusions from a study of representative items
is called induction. In statistics, induction is a generalization of facts on the
assumption that the results provided by an adequate sample may be taken
as applicable to the whole. The fact that the characteristics of the sample
provide a fairly good idea about the population characteristics is borne
out by the theory of probability. Sampling is based on two fundamental
principles of statistics theory viz, (i) the Law of Statistical Regularity and
(ii) the Law of Inertia of Large Numbers.

THE LAW OF STATISTICAL REGULARITY

The Law of Statistical Regularity is derived from the mathematical

theory of probability. According to W.I.King, “the Law of Statistical
Regularity formulated in the mathematical theory of probability lays down
that a moderately large number of items chosen at random from a very
large group are almost sure to have the characteristics of the large group.”
For example, if we want to find out the average income of 10,000 people,
69
we take a sample of 100 people and find the average. Suppose another
person takes another sample of 100 people from the same population and
finds the average, the average income found out by both the persons will
have the least difference. On the other hand if the average income of the
same 10,000 people is found out by the census method, the result will be
more or less the same.

Characteristics
1. The item selected will represent the universe and the result is
generalized to universe as a whole.
2. Since sample size is large, it is representative of the universe.
3. There is a very remote chance of bias.

LAW OF INERTIA OF LARGE NUMBERS

The Law of inertia of Large Numbers is an immediate deduction

from the Principle of Statistical Regularity. Law of Inertia of Large Numbers
states, “Other things being equal, as the sample size increases, the results
tend to be more reliable and accurate.” This is based on the fact that the
behavior or a phenomenon en masse. I.e., on a large scale is generally
stable. It implies that the total change is likely to be very small, when a
large number or items are taken in a sample. The law will be true on an
average. If sufficient large samples are taken from the patent population,
the reverse movements of different parts in the same will offset by the
corresponding movements of some other parts.

Sampling Errors:

In a sample survey, since only a small portion of the population

is studied its results are bound to differ from the census results and thus,
have a certain amount of error. In statistics the word error is used to denote
the difference between the true value and the estimated or approximated
value. This error would always be there no matter that the sample is drawn
at random and that it is highly representative. This error is attributed
to fluctuations of sampling and is called sampling error. Sampling error
exist due to the fact that only a sub set of the population has been used
to estimate the population parameters and draw inferences about the
population. Thus, sampling error is present only in a sample survey and is
completely absent in census method.
70
Sampling errors occur primarily due to the following reasons:

1. Faulty selection of the sample:

Some of the bias is introduced by the use of defective sampling

technique for the selection of a sample e.g. Purposive or judgment sampling
in which the investigator deliberately selects a representative sample to
obtain certain results. This bias can be easily overcome by adopting the
technique of simple random sampling.

2. Substitution:

When difficulties arise in enumerating a particular sampling unit

included in the random sample, the investigators usually substitute a
convenient member of the population. This obviously leads to some bias
since the characteristics possessed by the substituted unit will usually
be different from those possessed by the unit originally included in the
sample.

3. Faulty demarcation of sampling units:

Bias due to defective demarcation of sampling units is particularly

significant in area surveys such as agricultural experiments in the field of
crop cutting surveys etc. In such surveys, while dealing with border line
cases, it depends more or less on the discretion of the investigator whether
to include them in the sample or not.

4. Error due to bias in the estimation method:

Sampling method consists in estimating the parameters of the

population by appropriate statistics computed from the sample. Improper
choice of the estimation techniques might introduce the error.

5. Variability of the population:

Sampling error also depends on the variability or heterogeneity of

the population to be sampled.
Sampling errors are of two types: Biased Errors and Unbiased Errors

71
Biased Errors:

The errors that occur due to a bias of prejudice on the part of the
informant or enumerator in selecting, estimating measuring instruments
are called biased errors. Suppose for example, the enumerator uses the
deliberate sampling method in the place of simple random sampling
method, then it is called biased errors. These errors are cumulative in
nature and increase when the sample size also increases. These errors arise
due to defect in the methods of collection of data, defect in the method of
organization of data and defect in the method of analysis of data.

Unbiased Errors:

Errors which occur in the normal course of investigation or

enumeration on account of chance are called unbiased errors. They may
arise accidentally without any bias or prejudice. These errors occur due to
faulty planning of statistical investigation.
To avoid these errors, the statistician must take proper precaution and
care in using the correct measuring instrument. He must see that the
enumerators are also not biased. Unbiased errors can be removed with the
proper planning of statistical investigations. Both these errors should be
avoided by the statisticians.

Reducing Sampling Errors:

Errors in sampling can be reduced if the size of sample is increased.

This is shown in the following diagram.

From the above diagram it is clear that when the size of the sample
increases, sampling error decreases. And by this process samples can be
made more representatives to the population.

Testing of hypothesis:

As a part of investigation, samples are drawn from the population

and results are derived to help in taking the decisions. But such decisions
involve an element of uncertainty causing wrong decisions. Hypothesis is
an assumption which may or may not be true about a population parameter.
For example, if we toss a coin 200 times, we may get 110 heads and 90 tails.
72
At this instance, we are interested in testing whether the coin is unbiased
or not.

Therefore, we may conduct a test to judge the significance of the

difference of sampling or otherwise. To carry out a test of significance, the
following procedure has to be followed:

1. Framing the Hypothesis:

To verify the assumption, which is based on sample study, we

collect data and find out the difference between the sample value and the
population value. If there is no difference found or the difference is very
small then the hypothetical value is correct. Generally two hypotheses
complementary to each offer are constructed, and if one is found correct,
the other is rejected.

(a) Null Hypothesis:

The random selection of the samples from the given population

makes the tests of significance valid for us. For applying any test of
significance we first set up a hypothesis- a definite statement about the
population parameter/s. Such a statistical hypothesis, which is under
test, is usually a hypothesis of no difference and hence is called null
hypothesis. It is usually denoted by Ho. In the words of Prof. R.A.Fisher
“Null Hypothesis is the hypothesis which is tested for possible rejection
under the assumption that it is true.”

(b) Alternative Hypothesis.

Any hypothesis which is complementary to the null hypothesis

is called an alternative hypothesis. It is usually denoted by H1. It is very
important to explicitly state the alternative hypothesis in respect of any
null hypothesis H0 because the acceptance or rejection of Ho is meaningful
only if it is being tested against an opposite hypothesis. For example, if we
want to test the null hypothesis that the population has a specified mean
µ0(say), i.e., H0:µ=µ then the alternative hypothesis could be:
(i) H1:µ≠µ0 (i.e., µ>µ0 or µ<µ0)
(ii) H1: µ>µ0 (iii) H1: µ<µ0

73
 The alternative hypothesis (i) is known as a two-tailed alternative
and the alternatives in (ii) and (iii) are known as right-tailed and left-
tailed alternatives. Accordingly, the corresponding tests of significance are
called two-tailed, right-tailed and left-tailed tests respectively.

The null hypothesis consists of only a single parameter value and is

usually simple while alternative hypothesis is usually composite.

Types Of Errors In Testing Of Hypothesis:

As stated earlier, the inductive inference consists in arriving at a

decision to accept or reject a null hypothesis (Ho) after inspecting only
a sample from it. As such an element of risk – the risk of taking wrong
decision is involved. In any test procedure, the four possible mutually
disjoint and exhaustive decisions are:

i. Reject Ho when actually it is not true i.e., when Ho is false.

ii. Accept Ho when it is true.
iii. Reject Ho when it is true.
iv. Accept Ho when it is false.

The decisions in (i) and (ii) are correct decisions while the decisions
in (iii) and (iv) are wrong decisions. These decisions may be expressed in
the following dichotomous table:

Decision from sample

Reject Ho Accept Ho
True State
Wrong
Ho True Correct
Type I Error

Ho False Wrong
Correct
(H1True) Type II Error.

Thus, in testing of hypothesis we are likely to commit two types of
errors. The error of rejecting Ho when Ho is true is known as Type I Error
and the error of accepting Ho when Ho is false is known as Type II Error.

74
 For example, in the Industrial Quality Control, while inspecting
the quality of a manufactured lot, the Inspector commits Type I Error
when he rejects a good lot and he commits Type II Error when he accepts
a bad lot.

SUMMARY

These days questionnaire method of data collection has become

very popular. It is a very powerful tool to collect required data in shortest
period of time and with little expense. It is scientific too. But drafting
of questionnaire is a very skilled and careful work. Therefore, there are
certain requirements and essentials which should be followed at the time
of framing the questionnaire. They include the following viz., (i) the
size of the questionnaire should be small, (ii) questions should be very
clear in understanding, (iii) questions should be put in a logical order,
(iv) questions should have simple meaning etc. Apart from this, multiple
choice questions should be asked. Questionnaire should be pre tested
before going for final data collection. Information supplied should be
cross checked for any false or insufficient information. After all these
formalities have been completed, a covering note should accompany the
questionnaire explaining various purposes, designs, units and incentives.

There are two ways of survey- census survey and Sample survey
through which data can be collected. Census survey means total
enumeration i.e., collecting data from each and every unit of the universe,
whereas sample survey concentrates on collecting data from a few units of
the universe selected scientifically for the purpose. Since census method is
more time taking, expensive and labor intensive, it becomes impractical to
depend on it. Therefore, sample survey is preferred which is scientific, less
expensive, less time taking and less labor intensive too.

The merits and demerits of the methods are:

Merits:

It reduces cost, saves time and is more reliable. It provides more

detailed information and is sometimes the only method to depend upon
for administrative convenience and scientifically.

75
Demerits:

Sometimes samples may not be representative and may give illusory

conclusions. There are lack of experts and sometimes it is more difficult
than the census method, since there might arise personal bias and the
determination of the size of the sample might be very difficult.

Apart from these, the essentials of sampling that are to be followed

are: samples must be representative, samples must be homogeneous and
the number of samples must be adequate. When a researcher resorts to
sampling, he intends to collect some data which would help him to draw
results and finally take a decision. When he takes a decision it’s on the
basis of hypothesis which is precisely an assumption and is prone to two
types of errors-Type I Error and Type II Error. When a researcher rejects a
correct hypothesis, he commits type I error and when he accepts a wrong
hypothesis he commits type II error. The researcher should try to avoid
both types of errors but committing type II error is more harmful than
type I error.

SELF ASSESMENT QUESTIONS (SAQS)

1. Explain questionnaire and examine its main characteristics.

(refer to the introduction part of the questionnaire section)
2. Explain main requirements of a good questionnaire.
(refer to the sub points from 1 to 11)
3. What is sampling? Explain its main merits and demerits.
(refer to the introduction and the following part of the lesson)
4. What are null and alternative hypothesis? Explain.
(refer the point framing the hypothesis)
5. What are type i error and type ii error?
(refer to types of error in hypothesis)

-----

76
 CHAPTER II

Lesson 3. Experiments

Lesson Outline

ӹ Procedures Adopted In Experiments

ӹ Meaning Of Experiments
ӹ Research Design In Case Of Hypothesis Testing Research Studies
ӹ Basic Principles In Experimental Designs
ӹ Prominent Experimental Designs

Learning Objectives

After reading this lesson you should be able to understand the

ӹ Nature and meaning of experiments

ӹ Kinds of experiments

Introduction

The meaning of experiment lies in the process of examining the

truth of a statistical hypothesis related to some research problem. For
example, a researcher can conduct an experiment to examine the newly
developed medicine. Experiment is of two types: absolute experiment and
comparative experiment. When a researcher wants to determine the impact
of a fertilizer on the yield of a crop it is a case of absolute experiment. On
the other hand, if he wants to determine the impact of one fertilizer as
compared to the impact of some other fertilizer, the experiment will then
be called as a comparative experiment. Normally, a researcher conducts a
comparative experiment when he talks of designs of experiments.

77
Research design can be of three types:

1. Research design in the case of descriptive and diagnostic research

studies,

2. Research design in the case of exploratory research studies, and

3. Research design in the case of hypothesis testing research studies.

Here we are mainly concerned with the third one which is Research design
in the case of hypothesis testing research studies.

Research design in the case of hypothesis testing research studies:

Hypothesis testing research studies are generally known as

experimental studies. This is a study where a researcher tests the
hypothesis of causal relationships between variables. This type of study
requires some procedures which will not only reduce bias and increase
reliability, but will also permit drawing inferences about causality.
Most of the times, experiments meet these requirements. Prof. Fisher is
considered as the pioneer of this type of studies (experimental studies). He
did pioneering work when he was working at Rothamsted Experimental
Station in England which was a centre for Agricultural Research. While
working there, Prof. Fisher found that by dividing plots into different
blocks and then by conducting experiments in each of these blocks
whatever information that were collected and inferences drawn from them
happened to be more reliable. This was where he was inspired to develop
certain experimental designs for testing hypotheses concerning scientific
investigations. Nowadays, the experimental design is used in researches
relating to almost every discipline of knowledge. Prof. Fisher laid three
principles of experimental designs:

1. The Principle of Replication

2. The Principle of Randomization and
3. The Principle of Local Cont

1. The Principle Of Replication:

According to this principle, the experiment should be repeated

more than once. Thus, each treatment is applied in many experimental
78
units instead of one. This way the statistical accuracy of the experiments is
increased. For example, suppose we are going to examine the effect of two
varieties of wheat. Accordingly, we divide the field into two parts and grow
one variety in one part and the other variety in the other. Then we compare
the yield of the two parts and draw conclusion on that basis. But if we are
to apply the principle of replication to this experiment, then we first divide
the field into several parts, grow one variety in half of these parts and the
other variety in the remaining parts. Then we collect the data of yield of
the two varieties and draw conclusion by comparing the same. The result
so obtained will be more reliable in comparison to the conclusion we draw
without applying the principle of replication. The entire experiment can
be repeated several times for better results.

2. The Principle Of Randomization:

When we conduct an experiment, the principle of randomization

provides us a protection against the effects of extraneous factors. This
means that this principle indicates that the researcher should design or
plan the experiment in such a way that the variations caused by extraneous
factors can all be combined under the general heading of ‘chance’. For
example, when a researcher grows one variety of wheat , say , in the first
half of the parts of a field and the other variety he grows in the other half,
then it is just possible that the soil fertility may be different in the first half
in comparison to the other half. If this is so the researcher’s result is not
realistic. In this situation, he may assign the variety of wheat to be grown in
different parts of the field on the basis of some random sampling technique
i.e., he may apply randomization principle and protect himself against
the effects of the extraneous factors. Therefore, by using the principle of
randomization, he can draw a better estimate of the experimental error.

3. The Principle Of Local Control:

This is another important principle of experimental designs. Under

this principle, the extraneous factor which is the known source of variability
is made to vary deliberately over as wide a range as necessary. This needs
to be done in such a way that the variability it causes can be measured and
hence eliminated from the experimental error. The experiment should be
planned in such a way that the researcher can perform a two-way analysis
of variance, in which the total variability of the data is divided into three
79
components attributed to treatments (varieties of wheat in this case), the
extraneous factor (soil fertility in this case) and experimental error. In
short, through the principle of local control we can eliminate the variability
due to extraneous factors from the experimental error.

Kinds Of Experimental Designs And Control

Experimental designs refer to the framework of structure of

an experiment and as such there are several experimental designs.
Generally, experimental designs are classified into two broad categories:
informal experimental designs and formal experimental designs.
Informal experimental designs are those designs that normally use a less
sophisticated form of analysis based on differences in the magnitudes,
whereas formal experimental designs offer relatively more control and use
precise statistical procedures for analysis. Important experimental designs
are discussed below:

1. Informal Experimental Designs:

(i) Before and after without control design

(ii) After only with control design
(iii) Before and after with control design

2. Formal Experimental Designs:

(i) Completely randomized design (generally called C.R design)

(ii) Randomized block design (generally called R.B design)
(iii) Latin square design (generally called L.S design)
(iv) Factorial designs.

(i) Before and after without control design:

In this design, a single test group or area is selected and the

dependent variable is measured before introduction of the treatment.
Then the treatment is introduced and the dependent variable is measured
again after the treatment has been introduced. The effect of the treatment
would be equal to the level of the phenomenon after the treatment minus
the level of the phenomenon before the treatment. Thus, the design can be
presented in the following manner:
80
 Test area Level of phenomenon Treatment Level of phenomenon
Before treatment(X) introduced after treatment(Y)

Treatment Effect =(Y)-(X)

The main difficulty of such a design is that with the passage of time
considerable extraneous variations may be there in its treatment effect.

(ii) After-only with control design:

Two groups or areas are selected in this design and the treatment
is introduced into the test area only. Then the dependent variable is
measured in both the areas at the same time. Treatment impact is assessed
by subtracting the value of the dependent variable in the control area from
its value in the test area. The design can be presented in the following
manner:

Test area: Treatment introduced Level of phenomenon

after Treatment (Y)

Control area: Level of phenomenon

Without treatment (Z)

Treatment Effect = (Y)-(Z)

The basic assumption in this type of design is that the two areas
are identical with respect to their behavior towards the phenomenon
considered. If this assumption is not true, there is the possibility of
extraneous variation entering into the treatment effect.

(iii) Before and after with control design:

In this design, two areas are selected and the dependent variable
is measured in both the areas for an identical time-period before the
treatment. Thereafter, the treatment is introduced into the test area only
81
and the dependent variable is measured in both for and identical time –
period after the introduction of the treatment. The effect of the treatment
is determined by subtracting the change in the dependent variable in the
control area from the change in the dependent variable in test area. This
design can be shown in the following way:

Time Period I Time Period II

Test area: Level of phenomenon Treatment Level of phenomenon

Before treatment (X) Introduced After treatment (Y)

Control area: Level of phenomenon Level of phenomenon
Without treatment Without treatment
(A) (Z)

Treatment Effect = (Y-X)-(Z-A)

This design is superior to the previous two designs because it

avoids extraneous variation resulting both from the passage of time and
from non-comparability of the rest and control areas. But at times, due to
lack of historical data time or a comparable control area, we should prefer
to select one of the first two informal designs stated above.

2. Formal Experimental Design

(i) Completely randomized design:

This design involves only two principles i.e., the principle of
replication and the principle of randomization of experimental designs.
Among all other designs this is the simpler and easier because it’s procedure
and analysis are simple. The important characteristic of this design is
that the subjects are randomly assigned to experimental treatments. For
example, if the researcher has 20 subjects and if he wishes to test 10 under
treatment A and 10 under treatment B, the randomization process gives
every possible group of 10 subjects selected from a set of 20 an equal
opportunity of being assigned to treatment A and treatment B. One way
analysis of variance (one way ANOVA) is used to analyze such a design.

82
(ii) Randomized block design:

R. B. Design is an improvement over the C.R. design. In the R .B.

Design, the principle of local control can be applied along with the other
two principles of experimental designs. In the R.B. design, subjects are
first divided into groups, known as blocks, such that within each group the
subjects are relatively homogenous in respect to some selected variable.
The number of subjects in a given block would be randomly assigned
to each treatment. Blocks are the levels at which we hold the extraneous
factor fixed, so that its contribution to the total variability of data can be
measured. The main feature of the R.B. design is that, in this, each treatment
appears the same number of times in each block. This design is analyzed
by the two-way analysis of variance (two-way ANOVA) technique.

(iii) Latin squares design:

The Latin squares design (L.S design) is an experimental design

which is very frequently used in agricultural research. Since agriculture
depends upon nature to a large extent, the condition of research and
investigation in agriculture is different than the other studies. For example,
an experiment has to be made through which the effects of fertilizers on
the yield of a certain crop, say wheat, are to be judged. In this situation, the
varying fertility of the soil in different blocks in which the experiment has
to be performed must be taken into consideration; otherwise the results
obtained may not be very dependable because the output happens to be
the effects of not only of fertilizers, but also of the effect of fertility of
soil. Similarly there may be the impact of varying seeds of the yield. In
order to overcome such difficulties, the L.S. design is used when there are
two major extraneous factors such as the varying soil fertility and varying
seeds. The Latin square design is such that each fertilizer will appear five
times but will be used only once in each row and in each column of the
design. In other words, in this design, the treatment is so allocated among
the plots that no treatment occurs more than once in any one row or any
one column. This experiment can be shown with the help of the following
diagram:

83
 FERTILITY LEVEL
I II III IV V
X1 A B C D E

X2 B C D E A

X3 C D E A B

X4 D E A B C

X5 E A B C D

From the above diagram, it is clear that in L.S. design the field
is divided into as many blocks as there are varieties of fertilizers. Then,
each block is again divided into as many parts as there are varieties of
fertilizers in such a way that each of the fertilizer variety is used in each of
the block only once. The analysis of L.S. design is very similar to the two-
way ANOVA technique.

(iv) Factorial design:

Factorial designs are used in experiments where the effects of

varying more than one factor are to be determined. These designs are
used more in economic and social matters where usually a large number
of factors affect a particular problem. Factorial designs are usually of two
types:

(i) Simple factorial designs and

(ii) Complex factorial designs.

(i) Simple factorial design:

In simple factorial design, the effects of varying two factors on the

dependent variable are considered but when an experiment is done with
more than two factors, complex factorial designs are used. Simple factorial
design is also termed as a ‘two-factor-factorial design,’ whereas complex
factorial design is known as ‘multi-factor-factorial design.

84
(ii) Complex factorial designs:

When the experiments with more than two factors at a time are
conducted, it involves the use of complex factorial designs. A design which
considers three or more independent variables simultaneously is called a
complex factorial design. In case of three factors with one experimental
variable, two treatments and two levels, complex factorial design will
contain a total of eight cells. This can be seen through the following
diagram:

2x2x2 COMPLEX FACTORIAL DESIGN

Experimental variable

Treatment A Treatment B
Control Control Control Control
variable 2 variable 2 variable 2 variable 2
Level I Level II Level III Level II
Level I
Cell 1 Cell 3 Cell 5 Cell 7
Control
variable 2
Cell 2 Cell 4 Cell 6 Cell 8
Level II

A pictorial presentation is given of the design shown above in the following:

Experimental Variable

Treatment Treatment
Control Variable II A B

Level II
Level I
Level I

Level II
Control Variable I

85
 The dotted line cell in this diagram corresponds to cell I of the above
stated 2x2x2 design and is for treatment A, level I of the control variable
1, and levelIi of the control variable 2. From this design, it is possible to
determine the main effects for three variables i.e., one experimental and
true control variables. The researcher can also determine the interaction
between each possible pair of variables (such interactions are called ‘first
order interactions’) and interaction between variable taken in triplets
(such interactions are called second order interactions). In case of a 2x2x2
design, the further given first order interactions are possible:

Experimental variable with control variable 1 (or EV x CV 1);
Experimental variable with control variable 2 (or EV x CV 2);
Control variable 1 with control variable 2 (or CV 1 x CV 2);

There will be one second order interaction as well in the given design (it is
between all the three variables i.e., EV x CV 1 x CV 2).

To determine the main effect for the experimental variable, the
researcher must necessarily compare the combined mean of data in cells 1,
2, 3 and 4 for Treatment A with the combined mean of data in cells 5,6,7
and 8 for Treatment B. In this way the main effect of experimental variable,
independent of control variable 1 and variable 2, is obtained. Similarly, the
main effect for control variable 1, independent experimental variable and
control variable 2, is obtained if we compare the combined mean of data in
cells 1, 3, 5 and 7 with the combined mean of data in cells 2, 4, 6 and 8 of
our 2x2x2 factorial design. On similar lines, one can determine the effect
of control variable 2 independent of experimental variable and control
variable 1, if the combined mean of data in cells 1,2,5 and 6 are compared
with the combined mean of data in cells 3,4,7 and 8.

To obtain the first order interaction, say, for EV x CV 1 in the above
stated design, the researcher must necessarily ignore control variable 2
for which purpose he may develop 2x2 design from the 2x2x2 design by
combining the data of the relevant cells of the latter design as has been
shown on next page:

86
 Experimental Variable

Treatment A Treatment B

Level I Cells 1,3 Cells 5,7

Control
Variable 1 Level II Cells 2,4 Cells 6,8

Similarly, the researcher can determine other first order interactions.
The analysis of the first order interaction in the manner described above is
essentially a simple factorial analysis as only two variables are considered
at a time and the remaining ones are ignored. But the analysis of the second
order interaction would not ignore one of the three independent variables
in case of a 2x2x2 design. The analysis would be termed as a complex
factorial analysis.

It may, however, be remembered that the complex factorial design
need not necessarily be of 2x2x2 type design, but can be generalized to
any number and combinations of experimental and control independent
variables. Of course, the greater the number of independent variables
included in a complex factorial design, the higher the order of the
interaction analysis possible. But the overall task goes on becoming more
and more complicated with the inclusion of more and more independent
variables in our design.

Factorial designs are used mainly because of the two advantages -

(i) They provide equivalent accuracy (as happens in the case of experiments
with only one factor) with less labour and as such are source of economy.
Using factorial designs, we can determine the effects of two (in simple
factorial design) or more (in case of complex factorial design) factors
(or variables) in one single experiment. (ii) they permit various other
comparisons of interest. For example, they give information about such
effects which cannot be obtained by treating one single factor at a time.
The determination of interaction effects is possible in case of factorial
designs.

87
Conclusion

There are several research designs and the researcher must decide
in advance of collection and analysis of data as to which design would prove
to be more appropriate for his research project. He must give due weight
to various points such as type of universe and it’s nature, the objective of
the study, the source list or the sampling frame, desired standard accuracy
and the like when taking a decision in respect of the design for his research
project.

SUMMARY

Experiment is the process of examining the truth of a statistical

hypothesis related to some research problem. There are two types of
experiments - absolute and comparative. There are three types of research
designs - research design for descriptive and diagnostic research, research
design for exploratory research studies and research design for hypothesis
testing. Prof. Fisher has laid three principles of experimental design. They
are Principle of Replication, Principle of Randomization and Principle of
Local Control. There are different kinds of experimental designs. Some of
them are Informal experimental design, After only with control design,
Formal experimental design, Completely randomized design, Randomized
block design, Latin square design and Factorial design.

SELF ASSESMENT QUESTIONS (SAQS)

1. Explain the meaning and types of experiment (Ref. Introduction and

types of research design next to introduction)

2. Explain informal designs. (Ref. i,ii,iii in informal experiment design

portion.)

3. Explain formal experimental design and control. (Ref. i,ii,iii,iv in

formal experiment design section)

4. Explain complex factorial design.

***

88
 CHAPTER II

Lesson 4. Observation

1 Steps In Observation
Meaning And Characteristics Of Observation
Types Of Observation
Stages Of Observation
Problems, Merits And Demerits

Introduction

Observation is a method that employs vision as its main means

of data collection. It implies the use of eyes rather than of ears and the
voice. It is accurate watching and noting of phenomena as they occur
with regard to the cause and effect or mutual relations. It is watching
other persons’ behavior as it actually happens without controlling it. For
example, watching bonded labourer’s life, or treatment of widows and
their drudgery at home, provide graphic description of their social life and
sufferings. Observation is also defined as “a planned methodical watching
that involves constraints to improve accuracy”.

89
CHARACTERISTICS OF OBSERVATION

Scientific observation differs from other methods of data collection

specifically in four ways: (i) observation is always direct while other
methods could be direct or indirect; (ii) field observation takes place in
a natural setting; (iii) observation tends to be less structured; and (iv) it
makes only the qualitative (and not the quantitative) study which aims at
discovering subjects’ experiences and how subjects make sense of them
(phenomenology) or how subjects understand their life (interpretivism).

Lofland (1955) has said that this method is more appropriate

for studying lifestyles or sub-cultures, practices, episodes, encounters,
relationships, groups, organizations, settlements and roles etc. Black and
Champion (1976) have given the following characteristics of observation:

ӹӹ Behavior is observed in natural surroundings.

ӹӹ It enables understanding significant events affecting social relations
of the participants.
ӹӹ It determines reality from the perspective of observed person himself.
ӹӹ It identifies regularities and recurrences in social life by comparing
data in our study with that of other studies.

Besides, four other characteristics are:

ӹӹ Observation involves some control pertaining to the observation

and to the means he uses to record data. However, such controls do
not exist for the setting or the subject population.
ӹӹ It is focused on hypotheses-free inquiry.
ӹӹ It avoids manipulations in the independent variable i.e., one that is
supposed to cause other variable(s) and is not caused by them.
ӹӹ Recording is not selective.

Since at times, observation technique is indistinguishable from

experiment technique, it is necessary to distinguish the two.

90
Observation involves few controls than the experiment technique.

1. The behaviour observed in observation is natural, whereas in

experiment it is not always so.
2. The behavior observed in experiment is more molecular (of a smaller
unit), while one in observation is molar.
3. In observation, fewer subjects are watched for long periods of time in
more varied circumstances than in experiment.
4. Training required in observation study is directed more towards
sensitizing the observer to the flow of events, whereas training in
experiments serves to sharpen the judgment of the subject.
5. In observational study, the behavior observed is more diffused.
Observational methods differ from one another along several variables
or dimensions.

Merits:

1. We get original data

2. We get more accurate and reliable data.
3. Satisfactory information can be extracted by the investigator
through indirect questions.
4. Data are homogeneous and comparable.
5. Additional information can be gathered.
6. Misinterpretation of questions can be avoided.

Demerits:

1. It is time consuming and costs more.

SELF ASSESSMENT QUESTIONS (SAQS)

1. What are the characteristics of observation?

2. How do you differentiate observation from experiment?

----

91
92
 CHAPTER– III

Statistical Analysis

CONTENTS

1. Probability
2. Probability Distribution
2.1 Binomial Distribution
2.2 Poisson Distribution
2.3 Normal Distribution
3. Testing of Hypothesis
3.1 Small Sample
3.2 Large Sample Test
4. Χ2 Test

1. PROBABILITY

If an experiment is repeated under essentially homogeneous and

similar conditions, two possible conclusions can be arrived. They are: the
results are unique and the outcome can be predictable and result is not
unique but may be one of the several possible outcomes. In this context,
it is better to understand various terms pertaining to probability before
examining the probability theory. The main terms are explained as follows:

(i) Random Experiment

An experiment which can be repeated under the same conditions

and the outcome cannot be predicted under any circumstances is known

93
as random experiment. For example: An unbiased coin is tossed. Here
we are not in a position to predict whether head or tail is going to occur.
Hence, this type of experiment is known as random experiment.

(ii) Sample Space

A set of possible outcomes of a random experiment is known as

sample space. For example, in the case of tossing of an unbiased coin twice,
the possible outcomes are HH, HT, TH and TT. This can be represented in
a sample space as S= (HH, HT, TH, TT).

(iii) An Event

Any possible outcomes of an experiment are known as an event. In

the case of tossing of an unbiased coin twice, HH is an event. An event
can be classified into two. They are: (a) Simple events, and (ii) Compound
events. Simple event is an event which has only one sample point in the
sample space. Compound event is an event which has more than one
sample point in the sample space. In the case of tossing of an unbiased
coin twice HH is a simple event and TH and TT are the compound events.

(iv) Complementary Event

A and A’ are the complementary event if A’ consists of all those

sample point which is not included in A. For instance, an unbiased
dice is thrown once. The probability of an odd number turns up are
complementary to an even number turns up. Here, it is worth mentioning
that the probability of sample space is always is equal to one. Hence, the P
(A’) = 1 - P (A).

(v) Mutually Exclusive Events

A and B are the two mutually exclusive events if the occurrence of

A precludes the occurrence of B. For example, in the case of tossing of
an unbiased coin once, the occurrence of head precludes the occurrence
of tail. Hence, head and tail are the mutually exclusive event in the case
of tossing of an unbiased coin once. If A and B are mutually exclusive
events, then the probability of occurrence of A or B is equal to sum of their
individual probabilities. Symbolically, it can be presented as:
94
 P (A U B) = P (A) + P (B)

If A and B is joint sets, then the addition theorem of probability can be

stated as:

P (A U B ) = P(A) + P(B) - P(AB)

(vi) Independent Event

A and B are the two independent event if the occurrence of A does

not influence the occurrence of B. In the case of tossing of an unbiased
coin twice, the occurrence of head in the first toss does not influence the
occurrence of head or tail in the toss. Hence, these two events are called
independent events. In the case of independent event, the multiplication
theorem can be stated as the probability of A and B is the product of their
individual probabilities. Symbolically, it can be presented as:-

P (A B) = P (A) * P (B)

Addition Theorem of Probability

Let A and B be the two mutually exclusive events, then the

probability of A or B is equal to the sum of their individual probabilities.
(for detail refer mutually exclusive events)

Multiplication Theorem of Probability

Let A and B be the two independent events, then the probability

of A and B is equal to the product of their individual probabilities. (for
details refer independent events)

Example: The odds that person X speaks the truth are 4:1 and the odds
that Y speaks the truth are 3:1. Find the probability that:-

i. Both of them speak the truth,

ii. Any one of them speak the truth and
iii. Truth may not be told.

95
Solution:

The probability of X speaks the truth = 1/5

The probability that X speaks lie = 4/5
The probability that Y speaks the truth = 1/4
The probability that Y speaks lie =¼

(i) Both of them speak truth = P(X) * P(Y)

= 1/5 * 1/4
= 1/20 (independent event)

(ii) Any one of them speak truth = P(X) + P(Y) - P(X*Y)

= 1/5 + 1/4 - 1/5*1/4
= 8/20
= 2/5 (not mutually exclusive events)

(iii)Truth may not be told

= 1 – p(any one of them speak truth)( complementary event)
= 1 – 2/5
= 3/5.

2. PROBABILITY DISTRIBUTION

If X is discrete random variable which takes the values of x1,

x2,x3….. X n and the corresponding probabilities are p1, p2, ……….pn,
then, X follows the probability distribution. The two main properties of
probability distribution are: (i) P(Xi) is always greater than or equal to
zero and less than or equal to one, and (ii) the summation of probability
distribution is always equal to one. For example, tossing of an unbiased
coin twice.
Then the probability distribution is:

X (probability of obtaining head): 0 1 2

P(xi) : ¼ ½ ¼

Expectation of probability

Let X be the discrete random variable which takes the value of x1,
x2,…… xn then the respective probability is p1, p2, ………… pn, then the
96
expectation of probability distribution is p1x1 + p2x2 + ………….. + pnxn.
In the above example, the expectation of probability distribution is (0* ¼
+1*1/2+2*¼) =1.

2.1 BINOMIAL DISTRIBUTION

The binomial distribution also known as ‘Bernoulli Distribution’ is

associated with the name of a Swiss mathematician, James Bernoulli who
is also known as Jacques or Jakon (1654 – 1705). Binomial distribution
is a probability distribution expressing the probability of one set of
dichotomous alternatives. It can be explained as follows:

i. If an experiment is repeated under the same conditions for a fixed

number of trials, say, n.

ii. In each trial, there are only two possible outcomes of the experiment.
Let us define it as “success” or “failure”. Then the sample space of
possible outcomes of each experiment is:

iii. S = [failure, success]

iv. The probability of a success denoted by p remains constant from trial

to trial and the probability of a failure denoted by q which is equal to
(1 – p).

v. The trials are independent in nature i.e., the outcomes of any trial
or sequence of trials do not affect the outcomes of subsequent trials.
Hence, the multiplication theorem of probability can be applied for the
occurrence of success and failure. Thus, the probability of success or
failure is p.q.

vi. Let us assume that we conduct an experiment in n times. Out of which

x times be the success and failure is (n-x) times. The occurrence of
success or failure in successive trials is mutually exclusive events.
Hence, we can apply addition theorem of probability.

vii. Based on the above two theorems, the probability of success or failure
is

P(X) = nCxpxqn-x

97
 n!
--------------- . Px qn-x
x ! (n – x) !

where p = probability of success in a single trail, q = 1 – p, n =

Number of trials and x = no. of successes in n trials.

Thus, for an event A with probability of occurrence p and non-
occurrence q, if n trials are made, probability distribution of the number
of occurrences of A will be as set. If we want to obtain the probable
frequencies of the various outcomes in n sets of N trials, the following
expression shall be used: N(p + q)n

N(p + q)n = Npn + nC1pn-1q + nC2pn-2q2 + ……+ nCrpn-rqr + ……qn.

The frequencies obtained by the above expansion are known as

expected or theoretical frequencies. On the other hand, the frequencies
actually obtained by making experiments are called actual or observed
frequencies. Generally, there is some difference between the observed and
expected frequencies but the difference becomes smaller and smaller as N
increases.

Obtaining Coefficient Of The Binomial Distribution:

The following rules may be considered for obtaining coefficients from the
binomial expansion:

i. The first term is qn.,

ii. The second term is nC1qn-1p,

iii. In each succeeding term the power of q is reduced by 1 and the power
of p is increased by 1.

iv. The coefficient of any term is found by multiplying the coefficient

of the preceding term by the power of q in that preceding term, and
dividing the products so obtained by one more than the power of p
in that proceeding term.

98
Thus, when we expand (q + p)n, we will obtain the following:-

(p + q)n = pn + nC1pn-1q + nC2pn-2q2 + ……+ nCrpn-rqr + ……qn.

Where, 1, nC1, nC2 ……. are called the binomial coefficient. Thus in
the expansion of (p + q)4 we will have (p + q)4 = p4 +4p3q +6p2q2 + 4p1q3
+ q4 and the coefficients will be 1, 4, 6, 4, 1.

From the above binomial expansion, the following general relationships

should be noted:

i. The number of terms in a binomial expansion is always n + 1,

ii. The exponents of p and q, for any single term, when added together,
always sum to n.

iii. The exponents of p are n, (n – 1), (n – 2),…….1, 0, respectively and

the exponents of q are 0, 1, 2, ……(n – 1), n, respectively.

iv. The coefficients for the n + 1 terms of the distribution are always
symmetrical in nature.

Properties Of Binomial Distribution

The main properties of binomial distribution are:-

i. The shape and location of binomial distribution changes as p changes
for a given n or as n changes for a given p. As p increases for a fixed
n, the binomial distribution shifts to the right.

ii. The mode of the binomial distribution is equal to the value of x

which has the largest probability. The mean and mode are equal if
np is an integer.

iii. As n increases for a fixed p, the binomial distribution moves to the

right, flattens and spreads out.

iv. The mean of the binomial distribution is np and it increases as n

increases with p held constant. For larger n there are more possible
outcomes of a binomial experiment and the probability associated
with any particular outcome becomes smaller.
99
 v. If n is larger and if neither p nor q is too close to zero, the binomial
distribution can be closely approximated by a normal distribution
with standardized variable given by z = (X – np) / √npq.

vi. The various constants of binomial distribution are:

Mean = np
Standard Deviation = √npq
µ1 = 0
µ2 = npq
µ3 = npq(q – p)
µ4 = 3n2p2q2 + npq(1 – 6pq).

(q – p)2
Skewness = ---------
npq

1 – 6pq
Kurtosis = 3 + ---------
npq
Illustrations:

A coin is tossed four times. What is the probability of obtaining

two or more heads?

Solution:

when a coin is tossed the probabilities of head and tail in case of an

unbiased coin are equal, i.e., p = q = ½

The various possibilities for all the events are the terms of the
expansion (q+p)4

(p + q)4 = p4 + 4p3q + 6p2q2 + 4p1q3 + q4

Therefore, the probability of obtaining 2 heads is

6p2q2 = 6 x (½)2(½)2
= 3/8

100
The probability of obtaining 3 heads is 6p3q1 = 4 x (½)3(½)1
= 1/4

The probability of obtaining 4 heads is (q)4 = (½)4

= 1/16

Therefore, the probability of obtaining 2 or more heads is

3 1 1 11
--- + --- + --- = -----
8 4 16 16

Illustration:

Assuming that half the population is vegetarian so that the chance of

an individual being a vegetarian is ½ and assuming that 100 investigations
can take sample of 10 individuals to verify whether they are vegetarians,
how many investigation would you expect to report that three people or
less were vegetarians?

Solution:

n= 10, p, i.e., probability of an individual being vegetarian = ½.q =1 – p= ½

Using binomial distribution, we have P(r) = ncr qn-rpr
Putting the various values, we have
1
10c r(½)r(½)10 – r = 10cr = (½)10 = --------10cr
1024
The probability that in a sample of 10, three or less people are vegetarian
shall be given by:

P(0) + p(1) + p(2) + p(3)

1
= --------- [10c0 + 10c1 + 10c2 + 10c3]
1024

101
 1 176 11
= --------- [ 1 + 10 + 45 + 120] = -------- = -----
1024 1024 64

Hence out of 1000 investigators, the number of investigators who will

11
Report 3 or less vegetarians in a sample of 10 is 1000 x --- = 172.
64
2.2 POISSON DISTRIBUTION

Poisson distribution was derived in 1837 by a French Mathematician

Simeon D Poisson (1731 – 1840). In binomial distribution, the values of p
and q and n are given. There is a certainty of the total number of events.
But there are cases where p is very small and n is very large and such
case is normally related to poisson distribution. For example, persons
killed in road accidents, the number of defective articles produced by a
quality machine. Poisson distribution may be obtained as a limiting case
of binomial probability distribution, under the following condition.

i. P, successes, approach zero (p 0)

ii. np = m is finite.

The poisson distribution of the probabilities of occurrence of various rare

events (successes) 0,1,2,…. Are given below:

Number of Success (X) Probabilities p(X)

0 e-m
1 me-m/1!
2 m2e-m/ 2!
r mre-m/ r!
n mne-m/ n!

Where, e = 2.718, and m = average number of occurrence of given

distribution.

102
The poisson distribution is a discrete distribution with a parameter m.
The various constants are:

i. Mean = m=p
ii. Standard Deviation = √m
iii. Skewness β1 = 1/m
iv. Kurtosis, β2 = 3 + 1/m
v. Variance = m

Illustration:

A book contains 100 misprints distributed randomly throughout

its 100 pages. What is the probability that a page observed at random
contains at least two misprints? Assume Poisson Distribution.

Solution:

Total number of misprints 100

M = ------------------------------- = ----- = 1
Total number of pages 100

Probability that a page contains at least two misprints:

P(r≥2) = 1 – [p(0) + p (1)]

mre-m
P(r) = --------
r!

10e-1 1 1
P(0) = ------ = e = ---- = ---------
-1

0! e 2.7183

11e-1 1 1
P(1) = ------ = e-1 = ---- = ---------
1! e 2.7183
103
 1 1
P(0) + p(1) = ----------- + ----------- = 0.736
2.718 2.718

P(r≥2) = 1 – [p(0) + p (1)] = 1-0.736 = 0.264

Illustration:

If the mean of a Poisson distribution is 16, find (1) S.D.(2) B1

(3) B2 (4) µ3 (5) µ4

Solution:

m = 16

1. S.D. = √m = √16 = 4
2. β1 = 1/m = 1/16 = 0.625
3. β2 = 3 + 1/m = 3 + 0.625 = 3.0625
4. µ3 = m = 16
5. µ4 = m + 3m2 = 16 + 3(16)2 = 784

2.3 NORMAL DISTRIBUTION

The normal distribution was first described by Abraham Demoivre

(1667-1754) as the limiting form of binomial model in 1733. Normal
distribution was rediscovered by Gauss in 1809 and by Laplace in 1812.
Both Gauss and Laplace were led to the distribution by their work on the
theory of errors of observations arising in physical measuring processes
particularly in astronomy.

The probability function of a Normal Distribution is defined as:

1 -(x - µ)2 / 2σ2

P(X) = ------------ e
σ√2π
104
Where, X = Values of the continuous random variable, µ = Mean of the
normal random variable, e = 2.7183, π = 3.1416

Relation Between Binomial, Poisson And Normal Distributions

Binomial, Poisson and Normal Distribution are closely related to

one other. When N is large while the probability P of the occurrence of
an event is close to zero so that q = (1-p) the binomial distribution is very
closely approximated by the Poisson distribution with m = np.
The Poisson distribution approaches a normal distribution with
standardized variable (x – m)/ √m as m increases to infinity.

Normal distribution and its properties

The important properties of the normal distribution are:-

1. The normal curve is “bell shaped” and symmetrical in nature. The

distribution of the frequencies on either side of the maximum ordinate
of the curve is similar with each other.

2. The maximum ordinate of the normal curve is at x = µ. Hence the

mean, median and mode of the normal distribution coincide.

3. It ranges between - ∞ to + ∞

4. The value of the maximum ordinate is 1/ σ√2π.

5. The points where the curve change from convex to concave or vice
versa is at X = µ ± σ.

6. The first and third quartiles are equidistant from median.

7. The area under the normal curve distribution are:

a. µ ± 1σ covers 68.27% area;

b. µ ± 2σ covers 95.45% area.
c. µ ± 3σ covers 99.73% area.

105
 68.27%

95.45%

99.73%

µ - 36 µ - 26 µ - 16 µ=0 µ + 16 µ + 26 µ + 36
-3 -2 -1 Z=0 +1 +2 +3

8. When µ = 0 and σ = 1, then the normal distribution will be a standard

normal curve. The probability function of standard normal curve is

1 -x2/2
P(X) = ------------ e
√2π

The following table gives the area under the normal probability curve for
some important value of Z.

Distance from the mean ordinate in Area under the curve

Terms of ± σ

Z = ± 0.6745 0.50
Z = ± 1.0 0.6826
Z = ± 1.96 0.95
Z = ± 2.00 0.9544
Z = ± 2.58 0.99
Z = ± 3.0 0.9973

9. All odd moments are equal to zero.

10. Skewness = 0 and Kurtosis = 3 in normal distribution.

Illustration:

Find the probability that the standard normal value lies between 0 and 1.5

106
 0.4332 (43.32%)

Z= 0 Z = 1.5

As the mean, Z = 0.

To find the area between Z = 0 and Z = 1.5, look the area between 0 and
1.5, from the table. It is 0.4332 (shaded area)

Illustration:

The results of a particular examination are given below in a summary

form:

Result Percentage of candidates

Passed with distinction 10

Passed 60
Failed 30

It is known that a candidate gets plucked if he obtains less than

40 marks, out of 100 while he must obtain at least 75 marks in order to
pass with distinction. Determine the mean and standard deviation of the
distribution of marks assuming this to be normal.

Solution:

30% students get marks less than 40.

40 – X
Z = ---------- = -0.52 (from the table)
σ
107
 30% 20% 40% 10%

40 – X = -0.52σ ----------- (i)

10% students get more than 75

40% area = 75 – X = 1.28 ------------ (ii)

= 75 – X = 1.28σ

Subtract (ii) from (i)

40 – X = -0.52 σ

75 – X = 1.28 σ
--------------------
-35 = -1.8 σ
35 = 1.8 σ
1.80 σ = 35

35
σ = ------- = 19.4
1.80

Mean
40 – X = -0.52 x (19.4)

-X = -40 – 10.09 = 50.09

108
Illustration:

The scores observed by candidate in a certain test are normally

distributed with mean 1000 and standard deviation 200. What per cent of
candidates receive scores (i) less than 800, (ii) between 800 and 1200? (the
area under the curve between Z = 0 and Z = 1 is 0.34134).

Solution:
X = 1000; σ = 200

X–X
Z = ----------
σ

i) For X = 800

800 – 1000
Z = ------------- = -1
200

Area between Z = -1 and Z = 0 is 0.34134

Area for Z = -1 = 0.5 – 0.34134 = 0.15866

Therefore, the percentage = 0.15866 x 100 = 15.86%

ii) when, X = 1200,

1200 – 1000
Z = -------------- = 1
200

Area between Z = 0 and Z = 1 is 0.34134

Area between X = 400 to X = 600

i.e.,
Z = -1 and Z = 1 is 0.34134 + 0.34134 = 0.6826 = 68.26%

109
 0.6826

0.1586

800 1000 1200

3. TESTING OF HYPOTHESIS

3.1 Test of Significance for Large Samples

The test of significance for the large samples can be explained by the
following assumptions:

i. The random sampling distribution of statistics is approximately

normal.
ii. Sampling values are sufficiently close to the population value and
can be used for the calculation of standard error of estimate.

1. The Standard Error Of Mean.

In the case of large samples, when we are testing the significance of

statistic, the concept of standard error is used. It measures only sampling
errors. Sampling errors are involved in estimating a population parameter
from a sample, instead of including all the essential information in the
population.

(i) when standard deviation of the population is known, the formula is

σp
S.E. X = ----
√n
Where,

S.E.X = The standard error of the mean, σp = Standard deviation of
the population, and n = Number of observations in the sample.

(ii) when standard deviation of population is not known, we have to use

the standard deviation of the sample in calculating standard error of mean.
110
The formula is

σ (sample)
S.E. X = ------------
√n

Where,

σ = Standard deviation of the sample, and n = Sample size

Illustration:

A sample of 100 students from Pondicherry University was taken
and their average was found to be 116 lbs with a standard deviation of 20
lbs. Could the mean weight of students in the population be 125 pounds?

Solution:

Let us take the hypothesis that there is no significant difference
between the sample mean and the hypothetical population mean.

σ 20 20
S.E. X = ---- = -------- = -------- = 2
√n √100 10

Difference 125 – 116 9

-------------- = ------------- = ------- = 4.5

S.E.X 2 2

Since, the difference is more than 2.58 S.E.(1% level) it could not
have arisen due to fluctuations of sampling. Hence the mean weight of
students in the population could not be 125 lbs.

3.2 Test Of Significance For Small Samples

If the sample size is less than 30, then those samples may be regarded
as small samples. As a rule, the methods and the theory of large samples
are not applicable to the small samples. The small samples are used in
111
testing a given hypothesis, to find out the observed values, which could
have arisen by sampling fluctuations from some values given in advance.
In a small sample, the investigator’s estimate will vary widely from sample
to sample. An inference drawn from a smaller sample result is less precise
than the inference drawn from a large sample result.

t-distribution will be employed, when the sample size is 30

or less and the population standard deviation is unknown.

The formula is

( X - µ)
t = ------- x √n
σ
where,
σ = √ ∑(X – X)2/n – 1

Illustration:

The following results are obtained from a sample of 20 boxes of

mangoes:

Mean weight of contents = 490gms,

Standard deviation of the weight = 9 gms.
Could the sample come from a population having a mean of 500
gms?

Solution:

Let us take the hypothesis that µ = 510 gms.

( X - µ)
t = ------- x √n
σ

X = 500; µ = 510; σ = 10; n = 20.

112
 500 – 510
t = ------------- x √20
10

Df = 20 – 1 = 19 = (10/9) √20 = (10/9) x 4.47 = 44.7/9 = 4.96

Df = 19, t0.01 = 3.25
The computed value is less than the table value. Hence, our null hypothesis
is accepted.

4. CHI-SQUARE TEST

F, t and Z tests are based on the assumption that the samples were
drawn from normally distributed populations. The testing procedure
requires assumption about the type of population or parameters, and these
tests are known as ‘parametric tests’.

There are many situations in which it is not possible to make any
rigid assumption about the distribution of the population from which
samples are being drawn. This limitation has led to the development of
a group of alternative techniques known as non-parametric tests. Chi-
square test of independence and goodness of fit is a prominent example of
the use of non-parametric tests.

Though non-parametric theory developed as early as the middle
of the nineteenth century, it was only after 1945 that non-parametric tests
came to be used widely in sociological and psychological research. The
main reasons for the increasing use of non-parametric tests in business
research are:-

i. These statistical tests are distribution-free

ii. They are usually computationally easier to handle and understand
than parametric tests; and
iii. They can be used with type of measurements that prohibit the use of
parametric tests.

The χ2 test is one of the simplest and most widely used non-
parametric tests in statistical work. It is defined as:

113
 ∑(O– E)2
Χ2 = ------------
E
Where

O = the observed frequencies, and E = the expected frequencies.
Steps:

The steps required to determine the value of χ2are:

(i) Calculate the expected frequencies. In general the expected

frequency for any cell can be calculated from the following equation:

R X C
E = ------------
N

Where
E = Expected frequency, R = row’s total of the respective cell,
C = column’s total of the respective cell and N = the total number of
observations.

(ii) Take the difference between observed and expected frequencies

and obtain the squares of these differences. Symbolically, it can be
represented as (O – E)2

(iii) Divide the values of (O – E)2 obtained in step (ii) by the respective
expected frequency and obtain the total, which can be symbolically
represented by ∑[(O – E)2/E]. This gives the value of χ2 which can range
from zero to infinity. If χ2 is zero it means that the observed and expected
frequencies completely coincide. The greater the discrepancy between the
observed and expected frequencies, the greater shall be the value of χ2.

The computed value of χ2 is compared with the table value of χ2 for

given degrees of freedom at a certain specified level of significance. If at
the stated level, the calculated value of χ2 is less than the table value, the
difference between theory and observation is not considered as significant.

114
The following observation may be made with regard to the χ2 distribution:-
i. The sum of the observed and expected frequencies is always zero.
Symbolically, ∑(O – E) = ∑O - ∑E =N–N=0

ii. The χ2 test depends only on the set of observed and expected frequencies
and on degrees of freedom v. It is a non-parametric test.

iii. χ2 distribution is a limiting approximation of the multinomial

distribution.

iv. Even though χ2 distribution is essentially a continuous distribution it

can be applied to discrete random variables whose frequencies can be
counted and tabulated with or without grouping.

The Chi-Square Distribution

For large sample sizes, the sampling distribution of χ2 can be closely

approximated by a continuous curve known as the Chi-square distribution.
The probability function of χ2 distribution is:

F(χ2) = C (χ2)(v/2 – 1)e – x2/2

Where
e = 2.71828, v = number of degrees of freedom, C = a constant
depending only on v.

The χ2 distribution has only one parameter, v, the number of

degrees of freedom. As in case of t-distribution there is a distribution
for each different number of degrees of freedom. For very small number
of degrees of freedom, the Chi-square distribution is severely skewed
to the right. As the number of degrees of freedom increases, the curve
rapidly becomes more symmetrical. For large values of v the Chi-square
distribution is closely approximated by the normal curve.

115
The following diagram gives χ2 distribution for 1, 5 and 10 degrees of
freedom:

F(x2)

v=1

v=5
v = 10

0 2 4 6 8 10 12 14 16 18 20 22 χ2
χ2 Distribution

It is clear from the given diagram that as the degrees of freedom

increase, the curve becomes more and more symmetric. The Chi-square
distribution is a probability distribution and the total area under the curve
in each Chi-square distribution is unity.

Properties of χ2 Distribution

The main properties of χ2 distribution are:-

(i) The mean of the χ2 distribution is equal to the number of degrees

of freedom,

i.e.,
X=v

(ii) The variance of the χ2 distribution is twice the degrees of

freedom, Variance = 2v

(iii) µ1 = 0,

(iv) µ2 = 2v,

116
 (v) µ3 = 8v,

(vi) µ4 = 48v + 12v2.

µ32 64v2 8
(vii) β1 = --- = ----- = --
µ22 8v3 v

µ4 48v + 12v2 12
(v) β1µ3 = ------ = --------------- = 3 + ---
µ 22 4v2 v

The table values of χ2 are available only up to 30 degrees of freedom.

For degrees of freedom greater than 30, the distribution of χ2 approximates
the normal distribution. For degrees of freedom greater than 30, the
approximation is acceptable close. The mean of the distribution √2χ2 is
√2v – 1, and the standard deviation is equal to 1. Thus the application of
the test is simple, for deviation of √2χ2 from √2v – 1 may be interpreted as
a normal deviate with units standard deviation. That is,

Z = √2χ2 - √ 2v – 1

Alternative Method Of Obtaining The Value of χ2

In a 2x2 table where the cell frequencies and marginal totals are as below:

a b (a+b)

c d (c+d)

(a+c) (b+d) N

N is the total frequency and ad the larger cross-product, the value

of χ2 can easily be obtained by the following formula:

117
 N (ad – bc)2
χ2 = --------------------------------- or
(a + c) (b + d) (c + d) (a + b)

With Yate’s corrections

N (ab – bc - ½N)2
χ2 = -----------------------------------
(a + c) (b + d) (c + d) (a + b)

Conditions for Applying χ2 Test:

The main conditions considered for employing the χ2 test are:

(i) N must be to ensure the similarity between theoretically

correct distribution and our sampling distribution of χ2.

(ii) No theoretical cell frequency should be small when the expected

frequencies are too small. If it is so, then the value of χ2 will be overestimated
and will result in too many rejections of the null hypothesis. To avoid
making incorrect inferences, a general rule is followed that expected
frequency of less than 5 in one cell of a contingency table is too small to use.
When the table contains more than one cell with an expected frequency of
less than 5 then add with the preceding or succeeding frequency so that the
resulting sum is 5 or more. However, in doing so, we reduce the number of
categories of data and will gain less information from contingency table.

(iii) The constraints on the cell frequencies if any should be linear, i.e.,
they should not involve square and higher powers of the frequencies such
as ∑O = ∑E = N.

Uses of χ2 test:

The main uses of χ2 test are:

i. χ2 test as a test of independence. With the help of χ2 test, we can find

out whether two or more attributes are associated or not. Let’s assume
that we have n observations classified according to some attributes.

118
 We may ask whether the attributes are related or independent. Thus,
we can find out whether there is any association between skin colour
of husband and wife. To examine the attributes that are associated,
we formulate the null hypothesis that there is no association against
an alternative hypothesis and that there is an association between the
attributes under study. If the calculated value of χ2 is less than the
table value at a certain level of significance, we say that the result of the
experiment provides no evidence for doubting the hypothesis. On the
other hand, if the calculated value of χ2 is greater than the table value
at a certain level of significance, the results of the experiment do not
support the hypothesis.

ii. χ2 test as a test of goodness of fit. This is due to the fact that it enables
us to ascertain how appropriately the theoretical distributions such as
binomial, Poisson, Normal, etc., fit empirical distributions. When an
ideal frequency curve whether normal or some other type is fitted to
the data, we are interested in finding out how well this curve fits with
the observed facts. A test of the concordance of the two can be made
just by inspection, but such a test is obviously inadequate. Precision
can be secured by applying the χ2 test.

iii. χ2 test as a test of homogeneity. The χ2 test of homogeneity is an

extension of the chi-square test of independence. Tests of homogeneity
are designed to determine whether two or more independent random
samples are drawn from the same population or from different
populations. Instead of one sample as we use with independence
problem we shall now have 2 or more samples. For example, we may be
interested in finding out whether or not university students of various
levels, i.e., middle and richer poor income groups are homogeneous in
performance in the examination.

Illustration:

In an anti-diabetes campaign in a certain area, a particular

medicine, say x was administered to 812 persons out of a total population
of 3248. The number of diabetes cases is shown below:

119
 Treatment Diabetes No Diabetes Total

Medicine x 20 792 812

No Medicine x 220 2216 2436

Total 240 3008 3248

Discuss the usefulness of medicine x in checking malaria.

Solution:

Let us take the hypothesis that quinine is not effective in checking

diabetes. Applying χ2 test :

(A) X (B) 240 x 812
Expectation of (AB) = ------------ = ------------ = 60
N 3248

Or E1, i.e., expected frequency corresponding to first row and first column
is 60. The bale of expected frequencies shall be:

60 752 812
180 2256 2436
240 3008 3248

O E (O – E)2 (O – E)2/E

20 60 1600 26.667
220 180 1600 8.889
792 752 1600 2.218
2216 2256 1600 0.709

[∑(O – E)2/E] = 38.593

120
 χ2 = [∑(O – E)2/E] = 38.593

V = (r – 1) (c – 1) = (2 – 1) (2 – 1) = 1

For
v = 1, χ20.05 = 3.84

The calculated value of χ2 is greater than the table value. The hypothesis
is rejected. Hence medicine x is useful in checking malaria.

Illustration:

In an experiment on immunization of cattle from tuberculosis the

following results were obtained:

Affected not affected

Inoculated 10 20
Not inoculated 15 5

Calculate χ2 and discuss the effect of vaccine in controlling susceptibility

to tuberculosis (5% value of χ2 for one degree of freedom = 3.84).

Solution:

Let us take the hypothesis that the vaccine is not effective in

controlling susceptibility to tuberculosis. Applying χ2 test:

N(ad – bc)2 50 (11x5 – 20x15)2

χ2 = -------------------------- = ------------------------ = 8.3

(a+b) (c+d)(a+c)(b+d) 30x20x25x25

Since the calculated value of χ2 is greater than the table value the hypothesis
is not true. We, therefore, conclude the vaccine is effective in controlling
susceptibility to tuberculosis.

***
121
122
 CHAPTER – IV

Statistical Applications

A BRIEF INTRODUCTION TO STATISTICAL APPLICATIONS

A manager in a business organization – whether in the top level,

or the middle level, or the bottom level - has to perform an important
role of decision making. For solving any organizational problem – which
most of the times happens to be complex in nature -, he has to identify
a set of alternatives, evaluate them and choose the best alternative. The
experience, expertise, rationality and wisdom gained by the manager over
a period of time will definitely stand in good stead in the evaluation of the
alternatives available at his disposal. He has to consider several factors,
sometimes singly and sometimes jointly, during the process of decision
making. He has to deal with the data of not only his organization but also
of other competing organizations.
It would be a challenging situation for a manager when he has
to face so many variables operating simultaneously, something internal
and something external. Among them, he has to identify the important
variables or the dominating factors and he should be able to distinguish
one factor from the other. He should be able to find which factors have
similar characteristics and which factors stand apart. He should be able to
know which factors have an inter play with each other and which factors
remain independent. It would be advantageous to him to know whether
there is any clear pattern followed by the variables under consideration.
At times he may be required to have a good idea of the values that the
variables would assume in future occasions. The task of a manager becomes
all the more difficult in view of the risks and uncertainties surrounding
the future events. It is imperative on the part of a manager to understand
the impact of various policies and programmes on the development of
the organization as well as the environment. Also he should be able to
understand the impact of several of the environmental factors on his
organization. Sometimes a manager has to take a single stage decision and
at times he is called for to take a multistage decision on the basis of various
factors operating in a situation.
123
 Statistical analysis is a tool for a manager in the process of decision
making by means of the data on hand. All managerial activities involve
an analysis of data. Statistical approach would enable a manager to have a
scientific guess of the future events also. Statistical methods are systematic
and built by several experts on firmly established theories and consequently
they would enable a manager to overcome the uncertainties associated with
future occasions. However, statistical tools have their shortcomings too.
The limitations do not reflect on the subject. Rather they shall be traced
to the methods of data collection and recording of data. Even with highly
sophisticated statistical methods, one may not arrive at valid conclusions
if the data collected are devoid of representative character.

In any practical problem, one has to see whether the assumptions

are reasonable or not, whether the data represents a wide spectrum,
whether the data is adequate, whether all the conditions for the statistical
tests have been fulfilled, etc. If one takes care of these aspects, it would be
possible to arrive at better alternatives and more reliable solutions, thereby
avoiding future shocks. While it is true that a statistical analysis, by itself,
cannot solve all the problems faced by an organization, it will definitely
enable a manager to comprehend the ground realities of the situation. It
will for sure provide a foresight in the identification of the crucial variables
and the key areas so that he can locate a set of possible solutions within his
ambit. A manager has to have a proper blend of the statistical theories and
practical wisdom and he shall always strive for a holistic approach to solve
any organizational problem. A manager has to provide some safe-guarding
measures against the limitations of the statistical tools. In the process he
will be able to draw valid inferences thereby providing a clue as to the
direction in which the organization shall move in future. He will be ably
guided by the statistical results in the formulation of appropriate strategies
for the organization. Further, he can prepare the organization to face the
possible problems of business fluctuations in future and minimize the
risks with the help of the early warning signals indicated by the relevant
statistical tools.

A marketing manager of a company or a manager in a service

organization will have occasions to come across the general public and
consumers with several social and psychological variables which are
difficult to be measured and quantified.

124
Depending on the situation and the requirement, a manager may have
to deal with the data of just one variable (univariate data), or data on
two variables (bivariate data) or data concerning several simultaneous
variables (multivariate data).

The unit on hand addresses itself to the role of a manager as a

decision maker with the help of data available with him. Different statistical
techniques which are suitable for different requirements are presented
in this unit in a simple style. A manager shall know the strengths and
weaknesses of various statistical tools. He shall know which statistical
tool would be the most appropriate in a particular context so that the
organization will derive the maximum benefit out of it.

The interpretation of the results from statistical analysis occupies

an important place. Statistics is concerned with the aggregates and
not just the individual data items or isolated measurements of certain
variables. Therefore the conclusions from a statistical study will be valid
for a majority of the objects and normal situations only. There are always
extreme cases in any problem and they have to be dealt with separately.
Statistical tools will enable a manager to identify such outliers (abnormal
cases or extreme variables) in a problem. A manager has to evaluate the
statistical inferences, interpret them in the proper context and apply them
in appropriate situations.

While in an actual research problem, one has to handle a large

quantum of data, it is not possible to treat such voluminous data by
a beginner in the subject. Keeping this point in mind, any numerical
example in the present unit is based on a few data items only. It would be
worthwhile to the budding managers to make a start in solving statistical
problems by practicing the ones furnished in this unit.

The candidates are suggested to use hand calculators for solving
statistical problems. There will be frequent occasions to use statistical
tables of f-values furnished in this unit. The candidates are suggested to
have with them a copy of the tables for easy, ready reference. The books
and articles listed under the references may be consulted for further study
or applications of statistical techniques in relevant research areas.

***
125
126
 CHAPTER IV

1. Correlation And Regression Analysis

The Concept Of Correlation

Determination Of Simple Correlation Coefficient
Properties Of Correlation Coefficient
The Concept Of Rank Correlation
Determination Of Rank Correlation Coefficient
The Concept Of Regression
The Principle Of Least Squares
Normal Equations
Determination Of Regression Equations

127
SIMPLE CORRELATION

Correlation

Correlation means the average relationship between two or more

variables. When changes in the values of a variable affect the values of
another variable, we say that there is a correlation between the two
variables. The two variables may move in the same direction or in opposite
directions. Simply because of the presence of correlation between two
variables, we cannot jump to the conclusion that there is a cause-effect
relationship between them. Sometimes, it may be due to chance also.

Simple correlation

We say that the correlation is simple if the comparison involves

two variables only.

TYPES OF CORRELATION

Positive correlation

If two variables x and y move in the same direction, we say that

there is a positive correlation between them. In this case, when the value
of one variable increases, the value of the other variable also increases and
when the value of one variable decreases, the value of the other variable
also decreases. Eg. The age and height of a child.

Negative correlation

If two variables x and y move in opposite directions, we say that

there is a negative correlation between them. i.e., when the value of one
variable increases, the value of the other variable decreases and vice versa.
Eg. The price and demand of a normal good.

128
The following diagrams illustrate positive and negative correlations
between x and y.

y y

x x
Positive Correlation Negative Correlation

Perfect Positive Correlation

If changes in two variables are in the same direction and the changes
are in equal proportion, we say that there is a perfect positive correlation
between them.

Perfect Negative Correlation

If changes in two variables are in opposite directions and the

absolute values of changes are in equal proportion, we say that there is a
perfect negative correlation between them.

y y

x x
Perfect Positive Correlation Perfect Negative Correlation
129
Zero Correlation

If there is no relationship between the two variables, then the

variables are said to be independent. In this case the correlation between
the two variables is zero.

x
Zero Correlation

Linear Correlation'

If the quantum of change in one variable always bears a constant

ratio to the quantum of change in the other variable, we say that the two
variables have a linear correlation between them.

Coefficient of Correlation

The coefficient of correlation between two variables X, Y is a

measure of the degree of association (i.e., strength of relationship) between
them. The coefficient of correlation is usually denoted by ‘r’.

Karl Pearson’s Coefficient Of Simple Correlation:

Let N denote the number of pairs of observations of two variables X and Y.

The correlation coefficient r between X and Y is defined by

r=
∑ XY − ( ∑ X ) ( ∑ Y )
N

N ∑ X − (∑ X ) N ∑Y − (∑Y )
2 2 2 2

130
This formula is suitable for solving problems with hand calculators. To
apply this formula, we have to calculate ∑ X,∑Y, ∑XY, ∑X2, ∑Y2.

Properties Of Correlation Coefficient

Let r denote the correlation coefficient between two variables. r≥ is

interpreted using the following properties:

1. The value of r ranges from – 1.0 to 0.0 or from 0.0 to 1.0

2. A value of r = 1.0 indicates that there exists perfect positive correlation
between the two variables.
3. A value of r = - 1.0 indicates that there exists perfect negative
correlation between the two variables.
4. A value r = 0.0 indicates zero correlation i.e., it shows that there is no
correlation at all between the two variables.
5. A positive value of r shows a positive correlation between the two
variables.
6. A negative value of r shows a negative correlation between the two
variables.
7. A value of r = 0.9 and above indicates a very high degree of positive
correlation between the two variables.
8. A value of - 0.9 ≥ r > - 1.0 shows a very high degree of negative
correlation between the two variables.
9. For a reasonably high degree of positive correlation, we require r to
be from 0.75 to 1.0.
10. A value of r from 0.6 to 0.75 may be taken as a moderate degree of
positive correlation.

Problem 1

The following are data on Advertising Expenditure (in Rupees Thousand)

and Sales (Rupees in lakhs) in a company.

Advertising Expenditure : 18 19 20 21 22 23
Sales : 17 17 18 19 19 19
Determine the correlation coefficient between them and interpret the
result.

131
Solution:

We have N = 6. Calculate ∑ X, ∑Y, ∑XY, ∑Y2, ∑Y2 as follows:

X Y XY X2 Y2

18 17 306 324 289

19 17 323 361 289
20 18 360 400 324
21 19 399 441 361
22 19 418 484 361
23 19 437 529 361

Total :123 109 2243 2539 1985

The correlation coefficient r between the two variables is calculated as

follows:

N ∑ XY − ( ∑ X )( ∑ Y )
r=
N ∑ X 2 − (∑ X ) N ∑Y 2 − (∑Y )
2 2

6 × 2243 − 123 ×109

r=
6 × 2539 − (123) 6 ×1985 − (109 )
2 2

= (13458 – 13407) / {√(15234- 15129) √(11910- 11881)}

=51/{√105 √29} = 51/ (10.247 x 5.365)
= 51/ 54.975
= 0.9277

Interpretation

The value of r is 0.92. It shows that there is a high, positive

correlation between the two variables ‘Advertising Expenditure’ and
‘Sales’. This provides a basis to consider some functional relationship
between them.

132
Problem 2

Consider the following data on two variables X and Y.

X : 12 14 18 23 24 27
Y : 18 13 12 30 25 10

Determine the correlation coefficient between the two variables and

interpret the result.

Solution:

we have N = 6. Calculate ∑ X, ∑Y, ∑XY, ∑X2, ∑Y2 as follows:

X Y XY X2 Y2
12 18 216 144 324
14 13 182 196 169
18 12 216 324 144
23 30 690 529 900
24 25 600 576 625
27 10 270 729 100

Total : 118 108 2174 2498 2262

The correlation coefficient between the two variables is r =

{6 x 2174 – (118 x 108)} / { √(6 x 2498 - 1182) √(6 x 2262 - 1082) }

= (13044 – 12744) / {√(14988- 13924) √(13572- 11664)}

=300 / {√1064 √1908} = 300 / (32.62 x 43.68)
= 300 / 1424.84
= 0.2105

Interpretation

The value of r is 0.21. Even though it is positive, the value of r is

very less. Hence we conclude that there is no correlation between the
two variables X and Y. Consequently we cannot construct any functional
relational relationship between them.
133
Problem 3
Consider the following data on supply and price. Determine the correlation
Coefficient between the two variables and interpret the result.

Supply : 11 13 17 18 22 24 26 28
Price : 25 32 26 25 20 17 11 10

Determine the correlation coefficient between the two variables and

interpret the result.

Solution:

We have N = 8. Take X = Supply and Y = Price.

Calculate ∑ X, ∑Y, ∑XY, ∑X2, ∑Y2 as follows:

X Y XY X2 Y2

11 25 275 121 625

13 32 416 169 1024
17 26 442 289 676
18 25 450 324 625
22 20 440 484 400
24 17 408 576 289
26 11 286 676 121
28 10 280 784 100

Total: 159 166 2997 3423 3860

The correlation coefficient between the two variables is r =

{8 x 2997 – (159 x 166)} / { √(8 x 3423 - 1592) √(8 x 3860 - 1662) }

= (23976 – 26394) / {√(27384- 25281) √(30880- 27566)}

= - 2418 / {√2103 √3314}
= - 2418 / (45.86 x 57.57)
= - 2418 / 2640.16
= - 0.9159

134
Interpretation

The value of r is - 0.92. The negative sign in r shows that the two
variables move in opposite directions. The absolute value of r is 0.92
which is very high. Therefore we conclude that there is high negative
correlation between the two variables ‘Supply’ and ‘Price’.

Problem 4
Consider the following data on income and savings in Rs. Thousand.

Income : 50 51 52 55 56 58 60 62 65 66
Savings : 10 11 13 14 15 15 16 16 17 17

Determine the correlation coefficient between the two variables and

interpret the result.

Solution:
We have N = 10. Take X = Income and Y = Savings.
Calculate ∑ X, ∑Y, ∑XY, ∑X2, ∑Y2 as follows:

X Y XY X2 Y2
50 10 500 2500 100
51 11 561 2601 121
52 13 676 2704 169
55 14 770 3025 196
56 15 840 3136 225
58 15 870 3364 225
60 16 960 3600 256
62 16 992 3844 256
65 17 1105 4225 289
66 17 1122 4356 289
Total: 575 144 8396 33355 2126

The correlation coefficient between the two variables is r =

{10 x 8396 – (575 x 144)} / {√(10 x 33355 - 5752) √(10 x 2126 - 1442)}
= (83960 – 82800) / {√(333550- 330625) √(21260- 20736)}
= 1160 / {√2925 √524}
= 1160 / (54.08 x 22.89)
= 1160 / 1237.89 = 0.9371135
Interpretation

The value of r is 0.93. The positive sign in r shows that the

two variables move in the same direction. The value of r is very high.
Therefore we conclude that there is high positive correlation between
the two variables ‘Income’ and ‘Savings’. As a result, we can construct a
functional relationship between them.

RANK CORRELATION

Spearman’s rank correlation coefficient

If ranks can be assigned to pairs of observations for two variables X

and Y, then the correlation between the ranks is called the rank correlation
coefficient. It is usually denoted by the symbol ρ (rho). It is given by the
formula

6∑ D 2
ρ = 1−
N3 − N

where
D = difference between the corresponding ranks of X and Y

= RX − RY

and N is the total number of pairs of observations of X and Y.

136
Problem 5

Alpha Recruiting Agency short listed 10 candidates for final

selection. They were examined in written and oral communication skills.
They were ranked as follows:

Candidate’s
1 2 3 4 5 6 7 8 9 10
Serial no.
Rank in
written 8 7 2 10 3 5 1 9 6 4
communication
Rank
in oral 10 7 2 6 5 4 1 9 8 3
communication

Find out whether there is any correlation between the written and oral
communication skills of the short listed candidates.

Solution:

Take X = Written Communication Skill and Y = Oral Communication

Skill.

RANK OF X: R1 RANK OF Y: R2 D=R1- R2 D2

8 10 - 2 4
7 7 0 0
2 2 0 0
10 6 4 16
3 5 - 2 4
5 4 1 1
1 1 0 0
9 9 0 0
6 8 -2 4
4 3 1 1
Total: 30

We have N = 10. The rank correlation coefficient is

ρ = 1 - {6 ∑ D2 / (N3 – N)}
= 1 – {6 x 30 / (1000 – 10)}
137
 = 1 – (180 / 990)
= 1 – 0.18
= 0.82
Inference:

From the value of r, it is inferred that there is a high, positive rank

correlation between the written and oral communication skills of the short
listed candidates.

Problem 6

The following are the ranks obtained by 10 workers in abc company on the
basis of their length of service and efficiency.

Ranking as
1 2 3 4 5 6 7 8 9 10
per service
Rank as
2 3 6 5 1 10 7 9 8 4
perefficiency

Find out whether there is any correlation between the ranks obtained by
the workers as per the two criteria.

Solution:
Take X = Length of Service and Y = Efficiency.
Rank of X: R1 RANK OF Y: R2 D= R1- R2 D2

1 2 -1 1
2 3 -1 1
3 6 -3 9
4 5 -1 1
5 1 4 16
6 10 -4 16
7 7 0 0
8 9 -1 1
9 8 1 1
10 4 6 36

Total 82

138
We have N = 10. The rank correlation coefficient is
ρ = 1 - {6 ∑ D2 / (N3 – N)}
= 1 – { 6 x 82 / (1000 – 10) }
= 1 – (492 / 990)
= 1 – 0.497
= 0.503

Inference:
The rank correlation coefficient is not high.

Problem 7 (conversion of scores into ranks)

Calculate the rank correlation to determine the relationship

between equity shares and preference shares given by the following data
on their price.

Equity
90.0 92.4 98.5 98.3 95.4 91.3 98.0 92.0
share
Preference
76.0 74.2 75.0 77.4 78.3 78.8 73.2 76.5
share

Solution:

From the given data on share price, we have to find out the ranks for equity
shares and preference shares.

Step 1.
First, consider the equity shares and arrange them in descending
order of their price as 1,2,…,8. We have the following ranks:

Equity
98.5 98.3 98.0 95.4 92.4 92.0 91.3 90.0
share
Rank 1 2 3 4 5 6 7 8

Step 2.
Next, take the preference shares and arrange them in descending
order of their price as 1,2,…,8. We obtain the following ranks:

139
 Preference
78.8 78.3 77.4 76.5 76.0 75.0 74.2 73.2
share
Rank 1 2 3 4 5 6 7 8

Step 3.
Calculation of D2:

Fit the given data with the correct rank. Take X = Equity share and Y =
Preference share. We have the following table:

Rank of Rank of Y:
X Y D=R1- R2 D2
X: R1 R2
90.0 76.0 8 5 3 9

92.4 74.2 5 7 -2 4

98.5 75.0 1 6 -5 25

98.3 77.4 2 3 -1 1

95.4 78.3 4 2 2 4

91.3 78.8 7 1 6 36

98.0 73.2 3 8 -5 25

92.0 76.5 6 4 2 4

Total 108

Step 4.
Calculation of ρ:

We have N = 8. The rank correlation coefficient is

ρ = 1 - { 6 ∑ D2 / (N3 – N)}
= 1 – { 6 x 108 / (512 – 8) }
= 1 – (648 / 504)
= 1 – 1.29
= - 0.29
140
Inference:
From the value of ρ, it is inferred that the equity shares and
preference shares under consideration are negatively correlated. However,
the absolute value of ρ is 0.29 which is not even moderate.

Problem 8

Three managers evaluate the performance of 10 sales persons in an

organization and award ranks to them as follows:

Sales Person 1 2 3 4 5 6 7 8 9 10

Rank Awarded
8 7 6 1 5 9 10 2 3 4
by Manager I

Rank Awarded
7 8 4 6 5 10 9 3 2 1
by Manager II
Rank Awarded
by 4 5 1 8 9 10 6 7 3 2
Manager III

Determine which two managers have the nearest approach in the evaluation
of the performance of the sales persons.

Solution:

Manager Manager Manager

Sales
I II III (R1-R2)2 (R1-R3)2 (R2-R3)2
Person
Rank: R1 Rank:R2 Rank:R3
1 8 7 4 1 16 9

2 7 8 5 1 4 9
3 6 4 1 4 25 9
4 1 6 8 25 49 4
5 5 5 9 0 16 16
6 9 10 10 1 1 0
7 10 9 6 1 16 9

141
 8 2 3 7 1 25 16
9 3 2 3 1 0 1
10 4 1 2 9 4 1
Total 44 156 74

We have N = 10. The rank correlation coefficient between mangers I and

II is
ρ = 1 - { 6 ∑ D2 / (N3 – N)}
= 1 – { 6 x 44 / (1000 – 10) }
= 1 – (264 / 990)
= 1 – 0.27
= 0.73

The rank correlation coefficient between mangers I and III is

1 – { 6 x 156 / (1000 – 10) }

= 1 – (936 / 990)
= 1 – 0.95
= 0.05

The rank correlation coefficient between mangers II and III is

1 – { 6 x 74 / (1000 – 10) }
= 1 – (444 / 990)
= 1 – 0.44
= 0.56
Inference:

Comparing the 3 values of ρ, it is inferred that Mangers I and ii

have the nearest approach in the evaluation of the performance of the sales
persons.

Repeated values: Resolving ties in ranks

When ranks are awarded to candidates, it is possible that certain

candidates obtain equal ranks. For example, if two or three, or four
142
candidates secure equal ranks, a procedure that can be followed to resolve
the ties is described below.
We follow the Average Rank Method. If there are n items, arrange
them in ascending order or descending order and give ranks 1, 2, 3, …, n.
Then look at those items which have equal values. For such items, take the
average ranks.

If there are two items with equal values, their ranks will be two
consecutive integers, say s and s + 1. Their average is { s + (s+1)} / 2.
Assign this rank to both items. Note that we allow ranks to be fractions
also.

If there are three items with equal values, their ranks will be three
consecutive integers, say s, s + 1 and s + 2. Their average is { s + (s+1)
+ (s+2) } / 3 = (3s + 3) / 3 = s + 1. Assign this rank to all the three items.
A similar procedure is followed if four or more number of items has equal
values.

Correction term for ρ when ranks are tied

Consider the formula for rank correlation coefficient. We have

6∑ D 2
ρ = 1 −
N3 − N

If there is a tie involving m items, we have to add

m3 - m

12

to the term D2 in ρ. We have to add as many terms like (m3 – m) / 12 as

there are ties.

Let us calculate the correction terms for certain values of m. These are
provided in the following table.

143
 Correction term
m m 3
m -m
3

m3 - m
=
12

2 8 6 0.5

3 27 24 2

4 64 60 5

5 125 120 10

Illustrative examples:

If there is a tie involving 2 items, then the correction term is 0.5

If there are 2 ties involving 2 items each, then the correction term is
0.5 + 0.5 = 1

If there are 3 ties with 2 items each, then the correction term is
0.5 + 0.5 + 0.5 = 1.5

If there is a tie involving 3 items, then the correction term is 2

If there are 2 ties involving 3 items each, then the correction term is 2 + 2
=4

If there is a tie with 2 items and another tie with 3 items, then the correction
term is 0.5 + 2 = 2.5

If there are 2 ties with 2 items each and another tie with 3 items, then the
correction term is 0.5 + 0.5 + 2 = 3

144
Problem 9 : Resolving ties in ranks

The following are the details of ratings scored by two popular

insurance schemes. Determine the rank correlation coefficient between
them.

Scheme I 80 80 83 84 87 87 89 90
Scheme II 55 56 57 57 57 58 59 60

Solution:

From the given values, we have to determine the ranks.

Step 1.
Arrange the scores for Insurance Scheme I in descending order and
rank them as 1,2,3,…,8.

Scheme
I 90 89 87 87 84 83 80 80
Score

Rank 1 2 3 4 5 6 7 8

The score 87 appears twice. The corresponding ranks are 3, 4.

Their average is (3 + 4) / 2 = 3.5. Assign this rank to the two equal scores
in Scheme I.

The score 80 appears twice. The corresponding ranks are 7, 8.

Their average is (7 + 8) / 2 = 7.5. Assign this rank to the two equal scores
in Scheme I.

The revised ranks for Insurance Scheme I are as follows:

Scheme
I 90 89 87 87 84 83 80 80
Score

Rank 1 2 3.5 3.5 5 6 7.5 7.5

145
Step 2.
Arrange the scores for Insurance Scheme II in descending order
and rank them as 1,2,3,…,8.

Scheme
II 60 59 58 57 57 57 56 55
Score

Rank 1 2 3 4 5 6 7 8

The score 57 appears thrice. The corresponding ranks are 4, 5, 6.

Their average is (4 + 5 + 6) / 3 = 15 / 3 = 5. Assign this rank to the three
equal scores in Scheme II.

The revised ranks for Insurance Scheme II are as follows:

Scheme
60 59 58 57 57 57 56 55
II Score

Rank 1 2 3 5 5 5 7 8

Step 3.
Calculation of D2: Assign the revised ranks to the given pairs of
values and calculate D2 as follows:

Scheme I Scheme II Scheme I Scheme II

D=R1- R2 D2
Score Score Rank: R1 Rank: R2
80 55 7.5 8 - 0.5 0.25
80 56 7.5 7 0.5 0.25
83 57 6 5 1 1
84 57 5 5 0 0
87 57 3.5 5 - 1.5 2.25
87 58 3.5 3 0.5 0.25
89 59 2 2 0 0
90 60 1 1 0 0
Total 4

146
Step 4.
Calculation of ρ:

We have N = 8.

Since there are 2 ties with 2 items each and another tie with 3 items,
the correction term is 0.5 + 0.5 + 2 .

The rank correlation coefficient is

ρ = 1 - [{ 6 ∑ D2 + (1/2) + (1/2) +2 }/ (N3 – N)}]
= 1 – { 6 (4.+0.5+0.5+2) / (512 – 8) } = 1 – (6 x 7 / 504) = 1 - ( 42/504 )
= 1 - 0.083 = 0.917

Inference:

It is inferred that the two insurance schemes are highly, positively

correlated.

REGRESSION

In the pairs of observations, if there is a cause and effect relationship

between the variables X and Y, then the average relationship between
these two variables is called regression, which means “stepping back” or
“return to the average”. The linear relationship giving the best mean value
of a variable corresponding to the other variable is called a regression
line or line of the best fit. The regression of X on Y is different from the
regression of Y on X. Thus, there are two equations of regression and the
two regression lines are given as follows:

Regression of Y on X: Y −=
Y byx ( X − X )

Regression of X on Y: X − X= bxy (Y − Y )

Where X , Y are the means of X, Y respectively.

Result:
Let σx, σy denote the standard deviations of x, y respectively. We
have the following result.
147
 σ σX
=byx r=
Y
and bxy r
σX σY
= ∴ r 2 byx bxy=
and so r byx bxy

Result:
The coefficient of correlation r between X and Y is the square root
of the product of the b values in the two regression equations. We can find
r by this way also.

Application

The method of regression is very much useful for business

forecasting.

PRINCIPLE OF LEAST SQUARES

Let x, y be two variables under consideration. Out of them, let x

be an independent variable and let y be a dependent variable, depending
on x. We desire to build a functional relationship between them. For this
purpose, the first and foremost requirement is that x, y have a high degree
of correlation. If the correlation coefficient between x and y is moderate or
less, we shall not go ahead with the task of fitting a functional relationship
between them.

Suppose there is a high degree of correlation (positive or negative)

between x and y. Suppose it is required to build a linear relationship
between them i.e., we want a regression of y on x.

Geometrically speaking, if we plot the corresponding values of x

and y in a 2-dimensional plane and join such points, we shall obtain a
straight line. However, hardly we can expect all the pairs (x, y) to lie on
a straight line. We can consider several straight lines which are, to some
extent, near all the points (x, y). Consider one line. An observation (x1, y1)
may be either above the line of consideration or below the line. Project this
point on the x-axis. It will meet the straight line at the point (x1, y1e). Here
the theoretical value (or the expected value) of the variable is y1e while the
148
observed value is y1. When there is a difference between the expected and
observed values, there appears an error. This error is E1 = y1 –y1 . This is
positive if (x1, y1) is a point above the line and negative if (x1, y1) is a point
below the line. For the n pairs of observations, we have the following n
quantities of error:

E1 = y1 – y1 ,

E2 = y2 – y2 ,

En = yn – yn .

Some of these quantities are positive while the remaining ones are
negative. However, the squares of all these quantities are positive.

(X1, Y1)

e1
e2
(X2, Y2)

O
X

i.e.,
E21 = (y1 – y1 )2 ≥ 0, E22 = (y2 –y2)2 ≥ 0, …, E2n = (yn –yn )2 ≥ 0.

Hence the sum of squares of errors (SSE) = E21 + E22 + … + E2n

= (y1 –y2 )2 + (y2 –y2 )2 + … + (yn –yn)2 ≥ 0.

149
 Among all those straight lines which are somewhat near to the
given observations
(x1, y1), (x2, y2), …, (xn , yn) , we consider that straight line as the ideal one
for which the sse is the least. Since the ideal straight line giving regression
of y on x is based on this concept, we call this principle as the Principle of
least squares.

Normal equations

Suppose we have to fit a straight line to the n pairs of observations
(x1, y1), (x2, y2), …, (xn , yn). Suppose the equation of straight line finally
comes as
Y=a+bX (1)

Where
a, b are constants to be determined. Mathematically speaking, when
we require finding the equation of a straight line, two distinct points on
the straight line are sufficient. However, a different approach is followed
here. We want to include all the observations in our attempt to build a
straight line. Then all the n observed points (x, y) are required to satisfy
the relation

(1). Consider the summation of all such terms. We get

∑ y = ∑ (a + b x ) = ∑ (a .1 + b x ) = ( ∑ a.1) + ( ∑ b x ) = a ( ∑ 1 ) + b ( ∑ x).
i.e.
∑ y = an + b (∑ x) (2)

To find two quantities a and b, we require two equations. We have

obtained one equation i.e., (2). We need one more equation. For this
purpose, multiply both sides of (1) by

x. We obtain
x y = ax + bx2 .

Consider the summation of all such terms. We get

∑ x y = ∑ (ax + bx2 ) = (∑ a x) + ( ∑ bx2)

150
i.e.,
∑ x y = a (∑ x ) + b (∑ x2) ………….. (3)

Equations (2) and (3) are referred to as the normal equations associated
with the regression of y on x. Solving these two equations, we obtain

a =
∑ X ∑ Y - ∑ X ∑ XY
2

n ∑ X - (∑ X) 2 2

and b=
n ∑ XY - ∑ X ∑ Y
n ∑ X - (∑ X)
2 2

Note:
For calculating the coefficient of correlation,

we require ∑X, ∑Y, ∑ Xy, ∑ X2, ∑Y2.

For calculating the regression of y on x, we require ∑X, ∑Y, ∑ XY, ∑

X2. Thus, tabular column is same in both the cases with the difference that
∑Y2 is also required for the coefficient of correlation.

Next, if we consider the regression line of x on y, we get the equation

X = a + b y. The expressions for the coefficients can be got by interchanging
the roles of X and Y in the previous discussion. Thus, we obtain

a =
∑ Y ∑ X - ∑ Y ∑ XY
2

n ∑Y - (∑ Y) 2 2

And b =
n ∑ XY - ∑ X ∑ Y
n ∑Y - (∑ Y)
2 2

151
Problem 10

Consider the following data on sales and profit.

X 5 6 7 8 9 10 11

Y 2 4 5 5 3 8 7

Determine the regression of profit on sales.

Solution:

We have N = 7. Take X = Sales, Y = Profit.

Calculate ∑ X, ∑y, ∑XY, ∑X2 as follows:

X Y XY X2
5 2 10 25
6 4 24 36
7 5 35 49
8 5 40 64
9 3 27 81
10 8 80 100
11 7 77 121
Total: 56 34 293 476

a = {(∑ x2) (∑ y) – (∑ x) (∑ x y)} / {n (∑ x2) – (∑ x)2}

= (476 x 34 – 56 x 293) / ( 7 x 476 - 562 )

= (16184 – 16408 ) / ( 3332 – 3136 )

= - 224 / 196

= – 1.1429

152
 b = {n (∑ x y) – (∑ x) (∑ y)} / {n (∑ x2) – (∑ x) 2}

= (7 x 293 – 56 x 34)/ 196 = (2051 – 1904)/ 196

= 147 /196

= 0.75

The regression of Y on X is given by the equation

Y=a+bX
I.e.,
Y = – 1.14 + 0.75 X

Problem 11

The following are the details of income and expenditure of 10

households.

Income 40 70 50 60 80 50 90 40 60 60

Expenditure 25 60 45 50 45 20 55 30 35 30

Determine the regression of expenditure on income and estimate the

expenditure when the income is 65.

Solution:

We have N = 10. Take X = Income, Y = Expenditure

Calculate ∑ X, ∑y, ∑Xy, ∑X2 as follows:

X Y XY X2

40 25 1000 1600

70 60 4200 4900

50 45 2250 2500
153
 60 50 3000 3600

80 45 3600 6400

50 20 1000 2500

90 55 4950 8100

40 30 1200 1600

60 35 2100 3600
60 30 1800 3600

Total: 600 395 25100 38400

a = {(∑ x2) (∑ y) – (∑ x) (∑ x y)} / {n (∑ x2) – (∑ x) 2}

= ( 38400 x 395 - 600 x 25100 ) / (10 x 38400 - 6002)

= (15168000 – 15060000) / (384000 – 360000)

= 108000 / 24000

= 4.5

b = {n (∑ x y) – (∑ x) (∑ y)} / {n (∑ x2) – (∑ x) 2}

= ( 10 x 25100 – 600 x 395) / 24000

= (251000- 237000) / 24000

= 14000 / 24000

= 0.58

The regression of y on x is given by the equation

Y=a+bX
i.e.,
Y = 4.5 + 0.583 X
154
To estimate the expenditure when income is 65:

Take X = 65 in the above equation. Then we get

Y = 4.5 + 0.583 x 65

= 4.5 + 37.895

= 42.395

= 42 (approximately).

Problem 12

Consider the following data on occupancy rate and profit of a hotel.

Occupancy
40 45 70 60 70 75 70 80 95 90
rate

Profit 50 55 65 70 90 95 105 110 120 125

Determine the regressions of

(i) profit on occupancy rate and
(ii) occupancy rate on profit.

Solution:

We have N = 10. Take X = Occupancy Rate, Y = Profit.

Note that in Problems 10 and 11, we wanted only one regression

line and so we did not take ∑Y2 . Now we require two regression lines.
Therefore,

155
 Calculate ∑ X, ∑Y, ∑XY, ∑X2, ∑Y2.
X Y XY X2 Y2

40 50 2000 1600 2500

45 55 2475 2025 3025
70 65 4550 4900 4225
60 70 4200 3600 4900
70 90 6300 4900 8100
75 95 7125 5625 9025
70 105 7350 4900 11025
80 110 8800 6400 12100
95 120 11400 9025 14400
90 125 11250 8100 15625
Total: 695 885 65450 51075 84925

The regression line of Y on X:

Y=a+bX
Where
a ={(∑ x2) (∑ y) – (∑ x) (∑ x y)} / {n (∑ x2) – (∑ x) 2}
and
b ={n (∑ x y) – (∑ x) (∑ y)} / {n (∑ x2) – (∑ x) 2}

We obtain
a = (51075 x 885 – 695 x 65450) / (10x51075 - 6952)

= (45201375 – 45487750)/ (510750 – 483025)

= - 286375 / 27725

= - 10.329
b = (10 x 65450 – 695 x 885) / 27725

= (654500 – 615075) / 27725

= 39425 / 27725
= 1.422
156
So, the regression equation is Y = - 10.329 + 1.422 X

Next, if we consider the regression line of X on Y,

We get the equation X = a + b Y where

a = {(∑ y2) (∑ x) – (∑ y) (∑ x y)} / {n (∑ y2) – (∑ y) 2}

And
b = {n (∑ x y) – (∑ x) (∑ y)} / {n (∑ y2) – (∑ y) 2}.
We get

a = (84925 x 695 – 885 x 65450) / (10 x 84925 – 8852)

= (59022875 – 57923250) / ( 849250 – 783225)

= 1099625 / 66025

= 16.655,

b = (10 x 65450 – 695 x 885) / 66025

= (654500 – 615075) / 66025

= 39425 / 66025

= 0.597

So, the regression equation is X = 16.655 + 0.597 Y

Note:
For the data given in this problem, if we use the formula for r, we
get

r=
N∑ XY − ( ∑ X ) ( ∑ Y )
N ∑ X − (∑ X ) N ∑Y − (∑Y )
2 2 2 2

157
= (10 x 65450 – 695 x 885) / { √ (10 x 51075 - 6952 ) √ (10 x 84925 - 8852 ) }

= (654500 – 615075) / (√ 27725 √ 66025 )

= 39425 / 166.508 x 256.95

= 39425 / 42784.23

= 0.9214

However, once we know the two b values, we can find the coefficient of
correlation r between X and Y as the square root of the product of the
two b values.Thus we obtain

r = √ (1.422 x 0.597)
= √ 0.848934
= 0.9214.
Note that this agrees with the above value of r.

END OF CHAPTER QUESTIONS-SELF ASSESSMENT QUESTIONS(-

SAQ)

1. Explain the aim of ‘correlation analysis’.

2. Distinguish between positive and negative correlation.
3. State the formula for simple correlation coefficient.
4. State the properties of the correlation coefficient.
5. What is ‘rank correlation’? Explain.
6. State the formula for rank correlation coefficient.
7. Explain how to resolve ties while calculating ranks.
8. Explain the concept of regression.
9. What is the principle of least squares? Explain.
10. Explain normal equations in the context of regression analysis.
11. State the formulae for the constant term and coefficient in the regression
equation.
12. State the relationship between the regression coefficient and correlation
coefficient.
13. Explain the managerial uses of correlation analysis and regression
analysis.

***
158
 CHAPTER IV

2. Analysis Of Variance

Definition Of Anova
Assumptions Of Anova
Classification Of Linear Models
Anova For One-Way Classified Data
Anova Table For One-Way Classified Data
Null And Alternative Hypotheses
Type I Error
Level Of Significance
SS, Mss And Variance Ratio
Calculation Of F Value
Table Value Of F
Coding Method
Inference From Anova Table
Managerial Applications Of Anova

Understand the Concept of ANOVA

Formulate Null and Alternative Hypotheses
Construct ANOVA Table for One-Way Classified Data
Calculate T, N and CF
Calculate SS, DF and MSS
Calculate F Value
Find the Table Value of F
Draw Inference From ANOVA
Apply Coding Met
Understand the Managerial Applications of ANOVA

159
ANALYSIS OF VARIANCE (ANOVA)

Introduction

For managerial decision making, sometimes one has to carry out

tests of significance. The analysis of variance is an effective tool for this
purpose. The objective of the analysis of variance is to test the homogeneity
of the means of different samples.

Definition

According to R.A. Fisher, “analysis of variance is the separation of

variance ascribable to one group of causes from the variance ascribable to
other groups”.

Assumptions of ANOVA

The technique of ANOVA is mainly used for the analysis and

interpretation of data obtained from experiments. This technique is based
on three important assumptions, namely

1. The parent population is normal.

2. The error component is distributed normally with zero mean and

constant variance.

3. The various effects are additive in nature.

The technique of ANOVA essentially consists of partitioning the

total variation in an experiment into components of different sources of
variation. These sources of variations are due to controlled factors and
uncontrolled factors. Since the variation in the sample data is characterized
by means of many components of variation, it can be symbolically
represented in the mathematical form called a linear model for the sample
data.

160
Classification of models

Linear models for the sample data may broadly be classified into
three types as follows:

1. Random effect model

2. Fixed effect model
3. Mixed effect model

In any variance components model, the error component has

always random effects, since it occurs purely in a random manner. All
other components may be either mixed or random.

Random effect model

A model in which each of the factors has random effect (including

error effect) is called a random effect model or simply a random model.

Fixed effect model

A model in which each of the factors has fixed effects, buy only
the error effect is random is called a fixed effect model or simply a fixed
model.

Mixed effect model

A model in which some of the factors have fixed effects and some
others have random effects is called a mixed effect model or simply a
mixed model.

In what follows, we shall restrict ourselves to a fixed effect model.

In a fixed effect model, the main objective is to estimate the effects
and find the measure of variability among each of the factors and finally to
find the variability among the error effects.

The ANOVA technique is mainly based on the linear model which

depends on the types of data used in the linear model. There are several
types of data in ANOVA, depending on the number of sources of variation
namely,
161
 One-way classified data,

Two-way classified data, …

m-way classified data.

One-way classified data

When the set of observations is distributed over different levels of

a single factor, then it gives one-way classified data.

ANOVA for One-way classified data

Let y denote the jth observation corresponding to the ith level of

factor A and Yij the corresponding random variate.

Define the linear model for the sample data obtained from the
experiment by the equation

 i = 1, 2,..., k 
yij = µ + ai + eij  
 j = 1, 2,..., ni 

Where µ represents the general mean effect which is fixed and

which represents the general condition of the experimental units, ai
denotes the fixed effect due to ith level of the factor A (i=1,2,…,k) and
hence the variation due to ai (i=1,2,…,k) is said to be control.
The last component of the model eij is the random variable. It is called the
error component and it makes the Yij a random variate. The variation in eij
is due to all the uncontrolled factors and eij is independently, identically
and normally distributed with mean zero and constant variance σ2 .

For the realization of the random variate Yij, consider yij defined
by

 i = 1, 2,..., k 
yij = µ + ai + eij  
 j = 1, 2,..., ni 

162
 The expected value of the general observation yij in the experimental
units is given by
=
E ( yij ) µ=
i for all i 1, 2,..., k

With yij=µi+eij , where eij is the random error effect due to

uncontrolled factors (i.e., due to chance only).

Here we may expect µi=µ for all i=1,2,....,k , if there is no variation

due to control factors. If it is not the case, we have

µi ≠ µ for all i = 1, 2,..., k

i.e., µi − µ ≠ 0 for all i = 1, 2,..., k
Suppose µi − µ ≠ ai .
Then we have µi ≠ µ + ai for all i = 1, 2,..., k

On substitution for µi in the above equation, the linear model

reduces to

 i = 1, 2,..., k 
yij = µ + ai + eij   (1)
 j = 1, 2,..., ni 

The objective of ANOVA is to test the null hypothesis

H o : µi µ=
= for all i 1, 2,..., k =
or H o : ai 0=
for all i 1, 2,..., k . For
carrying out this test, we need to estimate the unknown parameters µ
ai for all i = 1, 2,..., k , by the principle of least squares. This can be done
by minimizing the residual sum of squares defined by

E = ∑ e 2ij
ij

= ∑(y
ij
ij − µ − ai ) 2 ,

Using (1). The normal equations can be obtained by partially

differentiating E with respect to µ and ai for all i = 1, 2,..., k and equating
the results to zero. We obtain

G N µ + ∑ ni ai (2)
=
i

163
and Ti = ni µ + ni ai, i = 1,2,…,k (3)

Where N = nk. We see that the number of variables (k+1) is more

than the number of independent equations (k). So, by the theorem on a
system of linear equations, it follows that unique solution for this system
is not possible.

However, by making the assumption that ∑n a

i
i i = 0 , we can get a

unique solution µ for and ai (i = 1,2,…,k). Using this condition in equation

(2), we get

G = Nµ
G
i.e. µ =
N

G
Therefore the estimate of µ is given by µµ= (4)
N

Again from equation (2), we have

T
i= µ + ai
ni

Ti
Hence, a=
i −µ
ni

Therefore, the estimate of ai is given by

Ti µ
aµ
=
i −µ
ni

i.e.,

Ti G
aµ
=
i − (5)
ni N

Substituting the least square estimates of µ and µi in the residual

sum of squares, we get

µ− $
E
= ∑(y
ij
ij −µ ai )2

164
After carrying out some calculations and using the normal equations (2)
and (3) we obtain

 G 2   Ti 2 G 2 
E=  ∑ yij2 − −∑ −  (6)
 ij N   i ni N 

The first term in the RHS of equation (6) is called the corrected

total sum of squares while ∑ yij2 is called the uncorrected total sum of
ij
squares.

for measuring the variation due to treatment (controlled factor),
we consider the null hypothesis that all the treatment effects are equal.

i.e.,

H o : µ1= µ2= ...= µk= µ

i.e., H o : µi = µ for all i = 1, 2,..., k

i.e., H o : µi − µ = 0 for all i = 1, 2,..., k
i.e., H o : ai = 0

Under H o , the linear model reduces to

 i = 1, 2,..., k 
yij= µ + eij  
 j = 1, 2,..., ni 

Proceeding as before, we get the residual sum of squares for this

hypothetical model as

  G2
= E1  ∑ yij2  − (7)
  N
ij

Actually, E1 contains the variation due to both treatment and

error. Therefore a measure of variation due to treatment can be obtained
by “ E1 − E ”. Using (6) and (7), we get
165
 k
Ti 2 G 2
E1 −=
E ∑
i =1 ni
−
N
(8)

The expression in (8) is usually called the corrected treatment sum

k
Ti 2
of squares while the term ∑
i =1 ni
is called uncorrected treatment sum

G2
of squares. Here it may be noted that is a correction factor (also
N

called a correction term). Since E is based on N-K free observations, it has

N - K degrees of freedom (df). Similarly, since E1 is based on N -1 free
observation, E1 has N -1 degrees of freedom. So E1 − E has K -1 degrees
of freedom.

When actually the null hypothesis is true, if we reject it on the basis

of the estimated value in our statistical analysis, we will be committing
Type – I Error. The probability for committing this error is referred to as
the denoted by α. The testing of the null hypothesis H o may be carried
out by F test. For given α, we have

Trss
TrMSS dF : F
=
F =
EMSS Ess k −1, N − k

dF

i.e., It follows F distribution with degrees of freedom K-1 and N-K.

All these values are represented in the form of a table called ANOVA
table, furnished below.

166
 ANOVA table for one-way classified data

Source Degrees Sum variance

Mean Squares
of of of ratio
(MS)
Variation Freedom Squares(SS) F

Between
the level of
the factor k-1 E1 − E =
QT QT MT
MT = FT = :
(Treat k −1 ME
k
Ti 2 G 2
ment) ∑i n − N Fk −1, N − k
i

QE
Within
the level of
N-k By subtrac QE -
factor ME =
tion N −k
(error)

Total N-1 - -
G2
=Q ∑ yij −
ij N

167
Variance ratio

The variance ratio is the ratio of the greater variance to the smaller
variance. It is also called the F-coefficient. We have

F = greater variance / smaller variance.

We refer to the table of F values at a desired level of significance

α. In general, α is taken to be 5 %. The table value is referred to as the
theoretical value or the expected value. The calculated value is referred to
as the observed value.

Inference

If the observed value of F is less than the expected value of F (i.e.,

Fo < Fe) for the given level of significance α , then the null hypothesis H o
is accepted. In this case, we conclude that there is no significant difference
between the treatment effects.

On the other hand, if the observed value of F is greater than the

expected value of F (i.e., ) for the given level of significance α , then the
null hypothesis H o is rejected. In this case, we conclude that all the
treatment effects are not equal.

Note:
If the calculated value of F and the table value of f are equal, we can
try some other value of α .

Problem 1

The following are the details of sales effected by three sales persons
in three door-to-door campaigns.

Sales person Sales in door – to – door campaign

A 8 9 5 10
B 7 6 6 9

C 6 6 7 5

168
 Construct an ANOVA table and find out whether there is any
significant difference in the performance of the sales persons.

Solution:

Method I (Direct method) :

∑ A = 8 + 9 + 5 + 10 = 32
∑ B = 7 + 6 + 6 + 9 = 28
∑ C = 6 + 6 + 7 + 5 = 24

32
A
Sample mean for A : = = 8
4

28
B
Sample mean for B : = = 7
4

24
C
Sample mean for C : = = 6
4

Total number of sample items = No. of items for A + No. of items

for B + No. of items for C

= 4 + 4 + 4 = 12

32 + 28 + 24 84
X
Mean of all the samples = = = 7
12 12

169
Sum of squares of deviations for A:

A A− A = A−8 ( A − A)
2

8 0 0

9 1 1

5 -3 9

10 2 4

14

Sum of squares of deviations for B:

B B− B = B−7 (B − B)
2

7 0 0

6 -1 1

6 -1 1
9 2 4
6

Sum of squares of deviations for C:

C C −C = C −6 (C − C )
2

6 0 0
6 0 0
7 -1 1
5 -1 1
2

170
Sum of squares of deviations within

Varieties = ∑ ( A − A ) + ∑ ( B − B ) + ∑ ( C − C )
2 2 2

= 14 + 6 + 2

= 22

Sum of squares of deviations for total variance:

Sales - X = sales
Sales person sales
( Sales − 7 )
2
–7

A 8 1 1

A 9 2 4

A 5 -2 4

A 10 3 9

B 7 0 0

B 6 -1 1

B 6 -1 1

B 9 2 4

C 6 -1 1

C 6 -1 1

C 7 0 0

C 5 2 4

30

171
ANOVA Table

Source of Degrees of Sum of squares

Variance
variation freedom of deviations

Between varieties 3–1 = 2 8 8

=4
2

Within varieties 12 – 3 = 9 22 22
= 2.44
9

Total 12 – 1 = 11 30

Calculation of F value:

Greater Variance 4.00

F = = = 1.6393
Smaller Variance 2.44

Degrees of freedom for greater variance ( df1 ) = 2Degrees of freedom for

smaller variance ( df1 ) = 9

Let us take the level of significance as 5%

The table value of F = 4.26

Inference:

The calculated value of F is less than the table value ofF. Therefore,
the null hypothesis is accepted. It is concluded that there is no significant
difference in the performance of the sales persons, at 5% level of
significance.

Method II (Short cut Method):

∑ A = 32, ∑ B = 28, ∑ C = 24.

T= Sum of all the sample items

172
 = ∑ A + ∑ B + ∑C
= 32 + 28 + 24
= 84

N = Total number of items in all the samples = 4 + 4 + 4 =12

T 2 842
Correction Factor = = = 588
N 12

Calculate the sum of squares of the observed values as follows:

Sales Person X X2

A 8 64
A 9 81
A 5 25
A 10 100
B 7 49
B 6 36
B 6 36

B 9 81

C 6 36

C 6 36

C 7 49

C 5 25
618

Sum of squares of deviations for total variance = ∑X 2

- correction
factor
= 618 – 588 = 30.

173
Sum of squares of deviations for variance between samples

( ∑ A) + ( ∑ B ) + ( ∑ C )
2 2 2

= − CF
N1 N2 N3
322 282 242
= + + − 588
4 4 4
1024 784 576
= + + − 588
4 4 4
= 256 + 196 + 144 − 588
=8

ANOVA Table

Source of Degrees of Sum of squares

Variance
variation freedom of deviations

Bet
3-1 = 2 8 8
ween varieties =4
2

Within varieties 12 – 3 = 9 22 22
= 2.44
9

Total 12 – 1 = 11 30

It is to be noted that the ANOVA tables in the methods I and II

are one and the same. For the further steps of calculation of F value and
drawing inference, refer to method I.

174
Problem 2

The following are the details of plinth areas of ownership apartment

flats offered by 3 housing companies A,B,C. Use analysis of variance to
determine whether there is any significant difference in the plinth areas of
the apartment flats.

H o u s i n g
Plinth area of apartment flats
Company
A 1500 1430 1550 1450

B 1450 1550 1600 1480

C 1550 1420 1450 1430

Use analysis of variance to determine whether there is any

significant difference in the plinth areas of the apartment’s flats.

Note:
As the given figures are large, working with them will be difficult.
Therefore, we use the following facts:

i). Variance ratio is independent of the change of origin.

ii.) Variance ratio is independent of the change of scale.

In the problem under consideration, the numbers vary from 1420

to 1600. So we follow a method called the coding method. First, let us
subtract 1400 from each item. We get the following transformed data:

Company Transformed measurement

A 100 30 150 50
B 50 150 100 80
C 150 20 50 30

Next, divide each entry by 10.

175
The transformed data are given below.

Company Transformed measurement

A 10 3 15 5

B 5 15 10 8

C 15 2 5 3

We work with these transformed data. We have

∑ A=10+3+15+5=33
∑ B =5+15+10+8=38
∑ C =15+2+5+3=25
∑T = ∑ A + ∑ B + ∑ C
= 33 + 38 + 25
= 96

N = Total number of items in all the samples = 4 + 4 + 4 = 12

T 2 962
Correction factor = = = 768
N 12

Calculate the sum of squares of the observed values as follows:

Company X X2
A 10 100
A 3 9
A 15 225
A 5 25
B 5 25
B 15 225
B 10 100
B 8 64
C 15 225

176
 C 2 4
C 5 25
C 3 9
1036

Sum of squares of deviations for total variance = ∑X 2

- correction
factor
= 1036 – 768 = 268

Sum of squares of deviations for variance between samples

( ∑ A) + ( ∑ B ) + ( ∑ C )
2 2 2

= − CF
N1 N2 N3
332 382 252
= + + − 768
4 4 4
1089 1444 625
= + + − 768
4 4 4
= 272.25 + 361 + 156.25 − 768
= 789.5 − 768
= 21.5

ANOVA Table

Source of Degrees of Sum of squares

Variance
variation freedom of deviation

Between varieties 3-1 = 2 21.5 21.5

= 10.75
2

Within varieties 12 – 3 = 9 264.5 24.65

= 27.38
9
Total 12 – 1 = 11 268

177
Calculation of F value:

Greater Variance 27.38

F= = = 2.5470
Smaller Variance 10.75

Degrees of freedom for greater variance ( df1 ) = 9

Degrees of freedom for smaller variance ( df 2 ) =2

The table value of f at 5% level of significance = 19.38

Inference:

Since the calculated value of F is less than the table value of F, the
null hypothesis is accepted and it is concluded that there is no significant
difference in the plinth areas of ownership apartment flats offered by the
three companies, at 5% level of significance.

Problem 3

A finance manager has collected the following information on the

performance of three financial schemes.

Sum of squares of
Source of variation Degrees of freedom
deviations
Treatments 5 15
Residual 2 25
Total (corrected) 7 40

Interpret the information obtained by him.

Note: ‘Treatments’ means ‘Between varieties’.

‘Residual’ means ‘Within varieties’ or ‘Error’.

178
Solution:

Number of schemes = 3 (since 3 – 1 = 2)

Total number of sample items = 8 (since 8 – 1 = 7)

Let us calculate the variance.

15
Variance between varieties = = 7.5
2

25
Variance between varieties = =5
5

Greater Variance 7.5

F = = = 1.5
Smaller Variance 5

Degrees of freedom for greater variance

( df1 ) = 2

Degrees of freedom for smaller variance

( df1 ) = 5

The total value of F at 5% level of significance

= 5.79

Inference:

Since the calculated value of F is less than the table value of F

we accept the null-hypothesis and conclude that there is no significant
difference in the performance of the three financial schemes.

179
SELF ASSESSMENT QUESTIONS---END OF CHAPTER
QUESTIONS.[IMPORTANT FOR EXAMS]

1. Define analysis of variance.

2. State the assumptions in analysis of variance.
3. Explain the classification of linear models for the sample
data.
4. Explain anova table.
5. Explain how inference is drawn from anova table.
6. Explain the managerial applications of analysis of variance.

***

180
 CHAPTER IV

3. Partial And Multiple Correlation

The Concept Of Partial Correlation

The Concept Of Multiple Correlation

Determine Partial Correlation Coefficient

Determine Multiple Correlation Coefficient

I. PARTIAL CORRELATION

Simple correlation is a measure of the relationship between a

dependent variable and another independent variable. For example, if
the performance of a sales person depends only on the training that he
has received, then the relationship between the training and the sales
performance is measured by the simple correlation coefficient r. However,
a dependent variable may depend on several variables. For example, the
yarn produced in a factory may depend on the efficiency of the machine,
the quality of cotton, the efficiency of workers, etc. It becomes necessary
to have a measure of relationship in such complex situations. Partial
correlation is used for this purpose. The technique of partial correlation
proves useful when one has to develop a model with 3 to 5 variables.

Suppose Y is a dependent variable, depending on n other variables

X1, X2, …, X n.. Partial correlation is a measure of the relationship between
Y and any one of the variables X1, X2,…,Xn, as if the other variables have
been eliminated from the situation.

181
The partial correlation coefficient is defined in terms of simple correlation
coefficients as follows:

Let r12. 3 denote the correlation of X1 and X2 by eliminating the

effect of X3.

Let r12 be the simple correlation coefficient between X1 and X2.

Let r13 be the simple correlation coefficient between X1 and X3.

Let r23 be the simple correlation coefficient between X2 and X3.

Then we have

r12 − r13 r 23
r12.3 =
(1 − r 2 13) (1 − r 2 23)

Similarly,

r13 − r12 r 32
r13.2 =
(1 − r 2 12) (1 − r 2 32)

r 23 − r 21 r13
and r 32.1 =
(1 − r 2 21) (1 − r 2 13)

182
Problem 1

Given that r12 = 0.6, r13 = 0.58, r23 = 0.70 determine the partial
correlation coefficient r12.3

Solution:

We have

0.6 − 0.58 x 0.70

=
(1 − (0.58)2 ) (1 − (0.70)2 )

0.6 − 0.406

=
(1 − 0.3364) (1 − 0.49)

0.194
=
0.6636 x 0.51

0.194

=
0.8146 x 0.7141

0.194
=
0.5817


= 0.3335

183
Problem 2

If r12 = 0.75, r13 = 0.80, r23 = 0.70, find the partial correlation
coefficient r13.2

Solution:

We have

r13 − r12 r 32
r13.2 =
(1 − r 2 12) (1 − r 2 32)

0.8 − 0.75 X 0.70

=
(1 − (0.75) 2 ) (1 − (0.70) 2 )

0.8 − 0.525
=
(1 − 0.5625) (1 − 0.49)

0.275
=
(0.4375) (0.51)

0.275
=
0.6614 X 0.7141

0.275
=
0.4723
= 0.5823

II. MULTIPLE CORRELATION

When the value of a variable is influenced by another variable, the

relationship between them is a simple correlation. In a real life situation, a
variable may be influenced by many other variables. For example, the sales
achieved for a product may depend on the income of the consumers, the
price, the quality of the product, sales promotion techniques, the channels
of distribution, etc. In this case, we have to consider the joint influence
184
of several independent variables on the dependent variable. Multiple
correlations arise in this context.

Suppose Y is a dependent variable, which is influenced by n other

variables X1, X2, …,Xn. The multiple correlation is a measure of the
relationship between Y and X1, X2,…, Xn considered together.

The multiple correlation coefficients are denoted by the letter R.

The dependent variable is denoted by X1. The independent variables are
denoted by X2, X3, X4,…, etc.

Meaning of notations:

R1.23 denotes the multiple correlation of the dependent variable

X1 with two independent variables X2 and X3 . It is a measure of the
relationship that X1 has with X2 and X3 .

R2.13 is the multiple correlation of the dependent variable X2 with

two independent variables X1 and X3.

R3.12 is the multiple correlation of the dependent variable X3 with

two independent variables X1 and X2.

R1.234 is the multiple correlation of the dependent variable X1 with

three independent variables X2 , X3 and X4.

Coefficient Of Multiple Linear Correlations

The coefficient of multiple linear correlation is given in terms of

the partial correlation coefficients as follows:

r 2 12 + r 2 13 - 2 r12 r13 r23

R1.23 =
1 - r 2 23

r 2 21 + r 2 23 - 2 r21 r23 r13

R 2.13 =
1 - r 2 13

185
 r 2 31 + r 2 32 - 2 r31 r32 r12
R 3.12 =
1 - r 2 12

Properties Of The Coefficient Of Multiple Linear Correlations:

1. The coefficient of multiple linear correlations R is a non-

negative quantity. It varies between 0 and 1.

2. R1.23 = R1.32

R2.13 = R2.31

R3.12 = R3.21, etc.

3. R1.23 ≥ |r12|,

R1.32 ≥ |r13|, etc.

Problem 3

If the simple correlation coefficients have the values r12 = 0.6, r13 =
0.65, r23 = 0.8, find the multiple correlation coefficient R1.23

Solution:

We have

r 2 12 + r 2 13 - 2 r12 r13 r23

R1.23 =
1 - r 2 23

(0.6) 2 + (0.65) 2 - 2x0.6x0.65x0.8

=
1 - (0.8) 2

186
 0.36+ 0.4225- 0.624
=
1 - 0.64

0.7825- 0.624
=
0.36

0.1585
=
0.36

= 0.4403

= 0.6636

Problem 4

Given that r21 = 0.7, r23 = 0.85 and r13 = 0.75, determine R2.13

Solution:

r 2 21 + r 2 23 - 2 r21 r23 r13

We have R 2.13 =
1 - r 2 13

(0.7) 2 + (0.85) 2 - 2 x0.7x0.85x0.75
=
1 - (0.75) 2

0.49+ 0.7225- 0.8925

=
1 - 0.5625

1.2125- 0.8925
=
0.4375

0.32
=
0.4375

= 0.7314
=0.8552
187
SELF ASSESSMENT QUESTIONS----END OF CHAPTER QUES-
TIONS[IMPORTANT FOR EXAMS]

1. Explain partial correlation.

2. Explain multiple correlations.
3. State the properties of the coefficient of multiple linear correla-
tions.

***

188
 CHAPTER IV

4. Factor Analysis And Conjoint Analysis

Factor Analysis
Conjoint Analysis
Steps In Development Of Conjoint Analysis
Applications Of Conjoint Analysis
Advantages And Disadvantages Of Conjoint Analysis
Illustrative Problems
Multi-Factor Evaluation Approach In Conjoint Analysis
Two-Factor Evaluation Approach In Conjoint Analysis

Understand The Concept Of Factor Analysis

Understand The Managerial Applications Of Factor Analysis
Understand The Concept Of Conjoint Analysis
Apply Rating Scale Technique In Conjoint Analysis
Apply Ranking Method In Conjoint Analysis
Apply Mini-Max Scaling Method In Conjoint Analysis
Understand Multi-Factor Evaluation Approach
Understand Two-Factor Evaluation Approach
Understand The Managerial Applications Of Conjoint Analy
sis

PART I - FACTOR ANALYSIS

In a real life situation, several variables are operating. Some variables

may be highly correlated among themselves. For example, if manager of
a restaurant has to analyse six attributes of a new product. He undertakes
a sample survey and finds out the responses of potential consumers. He

189
obtains the following attribute correlation matrix.

Attribute
1 2 3 4 5 6
1 1.00 0.05 0.10 0.95 0.20 0.02
2 0.05 1.00 0.15 0.10 0.60 0.85
3 0.10 0.15 1.00 0.50 0.55 0.10
4 0.95 0.10 0.50 1.00 0.12 0.08
5 0.20 0.60 0.55 0.12 1.00 0.80
6 0.02 0.85 0.10 0.08 0.80 1.00

Attribute Correlation Matrix

We try to group the attributes by their correlations. The high

correlation values are observed for the following attributes:

Attributes 1, 4 with a very high correlation coefficient of 0.95.

Attributes 2, 4 with a high correlation coefficient of 0.85.

Attributes 3, 4 with a high correlation coefficient of 0.85.

As a result, it is seen that not all the attributes are independent. The
attributes 1 and 4 have mutual influence on each other while the attributes
2, 5 and 6 have mutual influence among themselves. As far as attribute 3
is concerned, it has little correlation with the attributes 1, 2 and 6. Even
with the other attributes 4 and 5, its correlation is not high. However, we
can say that attribute 3 is somewhat closer to the variables 4 and 5 rather
than the attributes 1, 2 and 6. Thus, from the given list of 6 attributes, it is
possible to find out 2 or 3 common factors as follows:
I.
1) The common features of the attributes 1,3,4 will give a factor
2) The common features of the attributes 2, 5, 6 will give a factor

Or
II.
1) The common features of the attributes 1,4 will give a factor
2) The common features of the attributes 2,5,6 will give a factor
3) the attribute 3 can be considered to be an independent factor
190
 The factor analysis is a multivariate method. It is a statistical
technique to identify the underlying factors among a large number of
interdependent variables. It seeks to extract common factor variances from
a given set of observations. It splits a number of attributes or variables
into a smaller group of uncorrelated factors. It determines which variables
belong together. This method is suitable for the cases with a number of
variables having a high degree of correlation.

In the above example, we would like to filter down the attributes 1,

4 into a single attribute. Also we would like to do the same for the attributes
2, 5, 6. If a set of attributes (variables) A1, A2, …, Ak filter down to an
attribute Ai (1 ≤ i ≤ k), we say that these attributes are loaded on the
factor Ai or saturated with the factor Ai. Sometimes, more than one factor
also may be identified.

Basic concepts in factor analysis

The following are the key concepts on which factor analysis is

based.

Factor:

A factor plays a fundamental role among a set of attributes or
variables. These variables can be filtered down to the factor. A factor
represents the combined effect of a set of attributes. Either there may be
one such factor or several such factors in a real life problem based on the
complexity of the situation and the number of variables operating.

Factor loading:

A factor loading is a value that explains how closely the variables
are related to the factor. It is the correlation between the factor and the
variable. While interpreting a factor, the absolute value of the factor is
taken into account.

191
Communality:

It is a measure of how much each variable is accounted for by the

underlying factors together. It is the sum of the squares of the loadings of
the variable on the common factors. If A,B,C,… are the factors, then the
communality of a variable is computed using the relation

h2 = ( The factor loading of the variable with respect to factor A)2 +

( The factor loading of the variable with respect to factor B)2 +

( The factor loading of the variable with respect to factor C)2 + …..

Eigen value:

The sum of the squared values of factor loadings pertaining to a

factor is called an eigen value. It is a measure of the relative importance of
each factor under consideration.

Total Sum Of Squares (TSS)

It is the sum of the eigen values of all the factors.

Application Of Factor Analysis:

1. Model Building For New Product Development:

As pointed out earlier, a real life situation is highly complex and

it consists of several variables. A model for the real life situation can be
built by incorporating as many features of the situation as possible. But
then, with a multitude of features, it is very difficult to build such a highly
idealistic model. A practical way is to identify the important variables
and incorporate them in the model. Factor analysis seeks to identify
those variables which are highly correlated among themselves and find a
common factor which can be taken as a representative of those variables.
Based on the factor loading, some of variables can be merged together to
give a common factor and then a model can be built by incorporating such
factors. Identification of the most common features of a product preferred
by the consumers will be helpful in the development of new products.
192
2. Model Building For Consumers:

Another application of factor analysis is to carry out a similar

exercise for the respondents instead of the variables themselves. Using the
factor loading, the respondents in a research survey can be sorted out into
various groups in such a way that the respondents in a group have more
or less homogeneous opinions on the topics of the survey. Thus a model
can be constructed on the groups of consumers. The results emanating
from such an exercise will guide the management in evolving appropriate
strategies towards market segmentation.

PART II - CONJOINT ANALYSIS

Introduction

Everything in the world is undergoing a change. There is a proverb

saying that “the old order changes, yielding place to new”. Due to rapid
advancement in science and technology, there is fast communication across
the world. Consequently, the whole world has shrunk into something like
a village and thus now-a-days one speaks of the “global village”. Under the
present set-up, one can purchase any product of his choice from whatever
part of the world it may be available. Because of this reason, what was a
seller’s market a few years back has transformed into a buyer’s market now.

In a seller’s market of yesterday, the manufacturer or the seller could
pass on a product according to his own perceptions and prescriptions.
In the buyer’s market of today, a buyer decides what he should purchase,
what should be the quality of the product, how much to purchase, where
to purchase, when to purchase, at what cost to purchase, from whom
to purchase, etc. A manager is perplexed at the way a consumer takes
a decision on the purchase of a product. In this background, conjoint
analysis is an effective tool to understand a buyer’s preferences for a good
or service.

Meaning Of Conjoint Analysis

A product or service has several attributes. By an attribute, we

mean a characteristic, a property, a feature, a quality, a specification or
an aspect. A buyer’s decision to purchase a good or service is based on
193
not just one attribute but a combination of several attributes. i.e., he is
concerned with a join of attributes.

Therefore, finding out the consumer’s preferences for individual
attributes of a product or service may not yield accurate results for a
marketing research problem. In view of this fact, conjoint analysis seeks
to find out the consumer’s preferences for a ‘join of attributes’, i.e., a
combination of several attributes.

Let us consider an example. Suppose a consumer desires to purchase
a wrist watch. He would take into consideration several attributes of
a wrist watch, namely the configuration details such as mechanism,
size, dial, appearance, colour and other particulars such as strap, price,
durability, warranty, after-sales service, etc. If a consumer is asked what the
important aspect among the above list is, he would reply that all attributes
are important for him and so a manager cannot arrive at a decision on the
design of a wrist watch. Conjoint analysis assumes that the buyer will base
his decision not on just the individual attributes of the product but rather
he would consider various combinations of the attributes, such as

‘mechanism, colour, price, after-sales service’,

or ‘dial, colour, durability, warranty’,

or ‘dial, appearance, price, durability’, etc.

This analysis would enable a manager in his decision making

process in the identification of some of the preferred combinations of the
features of the product.

The rank correlation method seeks to assess the consumer’s

preferences for individual attributes. In contrast, the conjoint analysis
seeks to assess the consumer’s preferences for combinations (or groups)
of attributes of a product or a service. This method is also called an
‘unfolding technique’ because preferences on groups of attributes unfold
from the rankings expressed by the consumers. Another name for this
method is ‘multi-attribute compositional model’ because it deals with
combinations of attributes.

194
Steps In The Development Of Conjoint Analysis
The development of conjoint analysis comprises of the following
steps:

1. Collect a list of the attributes (features) of a product or a service.

2. For each attribute, fix a certain number of points or marks. The

more the number of points for an attribute, the more serious the consumers’
concern on that attribute.

3. Select a list of combinations of various attributes.

4. Decide a mode of presentation of the attributes to the respondents

of the study i.e., whether it should be in written form, or oral form, or a
pictorial representation etc.

5. Inform the combinations of the attributes to the prospective

customers.

6. Request the respondents to rank the combinations, or to rate them

on a suitable scale, or to choose between two different combinations at a
time.

7. Decide a procedure to aggregate the responses from the consumers.

Any one of the following procedures may be adopted:

i. Go by the individual responses of the consumers.

ii. Put all the responses together and construct a single utility
iii. Function.
iv. Split the responses into a certain number of segments such that within
each segment, the preferences would be similar.

8. Choose the appropriate technique to analyze the data collected

from the respondents.

9. Identify the most preferred combination of attributes.

10. Incorporate the result in designing a new product, construction of

an advertisement copy, etc.
195
Applications Of Conjoint Analysis

1. An idea of consumer’s preferences for combinations of attributes will

be useful in designing new products or modification of an existing
product.

2. A forecast of the profits to be earned by a product or a service.

3. A forecast of the market share for the company’s product.

4. A forecast of the shift in brand loyalty of the consumers.

5. A forecast of differences in responses of various segments of the

product.

6. Formulation of marketing strategies for the promotion of the product.

7. Evaluation of the impact of alternative advertising strategies.

8. A forecast of the consumers’ reaction to pricing policies.

9. A forecast of the consumers’ reaction on the channels of distribution.

10. Evolving an appropriate marketing mix.

11. Even though the technique of conjoint analysis was developed for the
formulation of corporate strategy, this method can be used to have
a comprehensive knowledge of a wide range of areas such as family
decision making process, pharmaceuticals, tourism development,
public transport system, etc.

Advantages Of Conjoint Analysis

1. The analysis can be carried out on physical variables.

2. Preferences by different individuals can be measured and pooled

together to arrive at a decision.

Disadvantages Of Conjoint Analysis

1. When more and more attributes of a product are included in the study,
the number of combinations of attributes also increases, rendering the
study highly difficult. Consequently, only a few selected attributes can
be included in the study.

196
2. Gathering of information from the respondents will be a tough job.

3. Whenever novel combinations of attributes are included, the

respondents will have difficulty in capturing such combinations.

4. The psychological measurements of the respondents may not be

accurate. In spite of the above stated disadvantages, conjoint analysis
offers more scope to the researchers in identifying the consumers’
preferences for groups of attributes.

Illustrative Problem 1 : Application Of Rating Scale Technique

A wrist watch manufacturer desires to find out the combinations

of attributes that a consumer would be interested in. After considering
several attributes, the manufacturer identifies the following combinations
of attributes for carrying out marketing research.

Combination – I Mechanism, colour, price, after-scales service

Combination – II Dial, colour, durability, warranty
Combination – III Dial, appearance, price, durability
Combination – IV Mechanism, dial, price, warranty

12 respondents are asked to rate the 4 combinations on the following

3-point rating scale.

Scale – 1 : Less important

Scale – 2 : Somewhat important
Scale – 3 : Very important

197
Their responses are given in the following table:

Rating of Combination
Res-
Combi- Combi- Combi- Combi-
pondent
nation I nation II nation III nation IV
No.
Some Some
Less Very
1 what what
important important
important important
Some Very Some
Less
2 what important what
important
important important
Some Some
Less Very
3 what what
important important
important important
Some
Less Less Very
4 what
important important important
important
Some
Very Very Less
5 what
important important important
important
Some Some
Very Less
6 what what
important important
important important
Some
Less Very Less
7 what
important important important
important
Some Some
Very Less
8 what what
important important
important important
Some Some
Very Less
9 what what
important important
important important
Some Some
Very Less
10 what what
important important
important important
Some Some
Very Very
11 what what
important important
important important

198
 Some
Very Less Very
12 what
important important important
important

Determine the most important and the least important combinations

of the attributes.

Solution:

Let us assign scores to the scales as follows:
Sl. No. Scale Score
1 Less important 1
2 Some what important 3
3 Very important 5

The scores for the four combinations are calculated as follows:

No. Of
Combi- Score for Total
Response Res
nation Response score
pondents

Less important 1 2 1X2=2

I Some what 3 6 3X6=18
important 5 4 5X4=20

12 40

Less important 1 5 1X5=5

II Some what 3 3 3X3=9
important 5 4 5X4=20

12 34

Less important 1 3 1X3= 3

III Some what 3 3 3X3= 9
important 5 6 5 X 6 = 30

12 42

199
Let us tabulate the scores earned by the four combinations as follows:

Combination Total scores

I 40

II 34

III 42

IV 32

Inference:

It is concluded that the consumers consider combination III as the

most important and combination IV as the least important.

Note:

For illustrating the concepts involved, we have taken up 12

respondents in the above problem. In actual research work, we should take
a large number of respondents, say 200 or 100. In any case, the number of
respondents shall not be less than 30.

Illustrative Problem 2: Application Of Ranking Method

A marketing manager selects four combinations of features of a

product for study. The following are the ranks awarded by 10 respondents.
Rank one means the most important and rank 4 means the least important.

Res-
pondent Rank Awarded
No.
Combination Combination Combination Combination
I II III IV
1 2 1 3 4
2 1 4 2 3

3 1 2 3 4

4 3 2 4 1

200
 5 4 1 2 3

6 1 2 3 4

7 4 3 2 1
8 3 1 2 4
9 3 1 4 2
10 4 1 2 3

Determine the most important and the least important combinations

of the features of the product.

Solution:

Let us assign scores to the ranks as follows:

Rank Score

1 10

2 8
3 6

4 4

The scores for the 4 combinations are calculated as follows:

No.
Com- Score
Rank of Total score
bination for rank
respondents
1 10 3 10 X 3 = 30
2 8 1 8 X 1= 8
I 3 6 3 6 X 3 = 18
4 4 3 4 X 3 = 12

10 68
1 10 5 10 X 5 = 50
2 8 3 8 X 3 = 24
II 3 6 1 6X1= 6
4 4 1 4X1= 4

201
 10 84
1 10 Nil --
2 8 5 8 X 5 = 40
III 3 6 3 6 X 3 = 18
4 4 2 4X2= 8

10 66

1 10 2 10 X 2 = 20
2 8 1 8X1= 8
IV 3 6 3 6 X 3 = 18
4 4 4 4 X 4 = 16

10 62

The final scores for the 4 combinations are as follows:

Combination Score

I 68

II 84

III 66

IV 62

Inference:

It is seen that combination II is the most preferred one by the

consumers and combination IV is the least preferred one.

Illustrative Problem 3:

Application Of Mini-Max Scaling Method

An insurance manager chooses 5 combinations of attributes of a

social security plan for analysis. He requests 10 respondents to indicate
their perceptions on the importance of the combinations by awarding
202
the minimum score and the maximum score for each combination in the
range of 0 to 100. The details of the responses are given below. Help
the manager in the identification of the most important and the least
important combinations of the attributes of the social security plan.

Com- Com- Com- Com- Com-

Res- bination bination bination bination bination
pondent I II III IV V
No.
Min Max Min Max Min Max Min Max Min Max
30 60 45 85 50 70 40 75 50 80
1
35 65 50 80 50 80 35 75 40 75
2
40 70 35 80 60 80 40 70 50 80
3
40 80 40 80 60 85 50 75 60 80
4
30 75 50 80 60 75 60 75 60 85
5
35 70 35 85 50 80 40 80 40 80
6
40 80 40 75 45 75 50 70 40 80
7
30 80 40 75 50 80 50 70 60 80
8
45 75 45 75 50 80 50 80 50 80
9
55 75 40 85 35 75 45 80 40 80
10

Solution:

For each combination, consider the minimum score and the

maximum score separately and calculate the average in each case.

Com- Com- Com- Com- Com-

bination bination bination bination bination
I II III IV V

Min Max Min Max Min Max Min Max Min Max

Total 380 730 420 800 510 780 460 750 490 800

Average 38 73 42 80 51 78 46 75 49 80

203
 Consider the mean values obtained for the minimum and maximum
of each combination and calculate the range for each combination as

Range = Maximum Value – Minimum Value

The measure of importance for each combination is calculated as follows:

Measure of importance for a combination of attributes

Range for that combination

= × 100
Sum of the ranges for all the combinations

Tabulate the results as follows:

Measure of
Combination Max. Value Min. Value Range
Importance

I 73 38 35 21.875
II 80 42 38 23.750
III 78 51 27 16.875
IV 75 46 29 18.125
V 80 49 31 19.375

Sum of the ranges 160 100.000

Inference:

It is concluded that combination II is the most important one and

combination III is the least important one.

APPROACHES FOR CONJOINT ANALYSIS

The following two approaches are available for conjoint analysis:

1. Multi-factor evaluation approach

2. Two-factor evaluation approach

204
MULTI-FACTOR EVALUATION APPROACH IN CONJOINT
ANALYSIS

Suppose a researcher has to analyze n factors. It is possible that

each factor can assume a value in different levels.

Product Profile

A product profile is a description of all the factors under

consideration, with any one level for each factor.

Suppose, for example, there are 3 factors with the levels given below.

Factor 1 : 3 levels
Factor 2 : 2 levels
Factor 3 : 4 levels

Then we have 3 × 2 × 4 =24 product profiles. For each respondent
in the research survey, we have to provide 24 data sheets such that each
data sheet contains a distinct profile. In each profile, the respondent is
requested to indicate his preference for that profile in a rating scale of 0
to 10. A rating of 10 indicates that the respondent’s preference for that
profile is the highest and a rating of 0 means that he is not all interested in
the product with that profile.

Example:
consider the product ‘Refrigerator’ with the following factors and
levels:

Factor 1 : Capacity of 180 liters; 200 liters; 230 liters

Factor 2 : Number of doors: either 1 or 2
Factor 3 : Price : rs. 9000; rs. 10,000; rs. 12,000

Sample profile of the product

Profile Number :
Capacity : 200 liters
Number of Doors : 1
205
 Price : Rs. 10,000
Rating of Respondent:
(in the scale of 0 to 10)

Steps In Multi-Factor Evaluation Approach:

1. Identify the factors or features of a product to be analyzed. If they are
too many, select the important ones by discussion with experts.

2. Find out the levels for each factor selected in step 1.

3. Design all possible product profiles. If there are n factors with levels
L1, L2,…Ln respectively, then the total number of profiles = L1L2…Ln.

4. Select the scaling technique to be adopted for multi-factor evaluation

approach (rating scale or ranking method).

5. Select the list of respondents using the standard sampling technique.

6. Request each respondent to give his rating scale for all the profiles of
the product. Another way of collecting the responses is to request each
respondent to award ranks to all the profiles: i.e., rank 1 for the best
profile, rank 2 for the next best profile etc.

7. For each factor profile, collect all the responses from all the participating
respondents in the survey work.

8. With the rating scale awarded by the respondents, find out the score
secured by each profile.
9. Tabulate the results in step 8. Select the profile with the highest score.
This is the most preferred profile.

10. implement the most preferred profile in the design of a new product.

Two-Factor Evaluation Approach In Conjoint Analysis

When several factors with different levels for each factor have to be
analyzed, the respondents will have difficulty in evaluating all the profiles
in the multi-factor evaluation approach. Because of this reason, two-factor
evaluation approach is widely used in conjoint analysis.

206
 Suppose there are several factors to be analyzed with different
levels of values for each factor, then we consider any two factors at a time
with their levels of values. For each such case, we have a data sheet called
a two-factor table. If there are n factors, then the number of such data
sheets is .

 n  n(n − 1)
  =
 2 2

Let us consider the example of ‘Refrigerator’ described in the multi-
factor approach. For the two factors (i) capacity and (ii) price, we have the
following data sheet.

Data Sheet (Two Factor Table) No:

Factor: price of refrigerator

Factor: Capacity Price

of Refrigerator Rs. 9,000 Rs. 10,000 Rs. 12,000

180 liters
200 liters
230 liters

In this case, the data sheet is a matrix of 3 rows and 3 columns.

Therefore, there are 3 × 3 = 9 places in the matrix. The respondent has
to award ranks from 1 to 9 in the cells of the matrix. A rank of 1 means
the respondent has the maximum preference for that entry and a rank of 9
means he has the least preference for that entry. Compared to multi-factor
evaluation approach, the respondents will find it easy to respond to two-
factor evaluation approach since only two factors are considered at a time.

Steps in two-factor evaluation approach:

Identify the factors or features of a product to be analyzed.

1. Find out the levels for each factor selected in step 1.

2. Consider all possible pairs of factors. If there are n factors, then the

207
  n  n(n − 1)
number of pairs is   = . For each pair of factors, prepare
 2 2

a two-factor table, indicating all the levels for the two factors. If L1 and L2
are the respective levels for the two factors, then the number of cells in the
corresponding table is L1L2.

3. Select the list of respondents using the standard sampling technique.

4. Request each respondent to award ranks for the cells in each two-
factor table. I.e., rank 1 for the best cell, rank 2 for the next best cell, etc.

5. For each two-factor table, collect all the responses from all the
participating respondents in the survey work.

6. With the ranks awarded by the respondents, find out the score
secured by each cell in each two-factor table.

7. Tabulate the results in step 7. Select the cell with the highest score.
Identify the two factors and their corresponding levels.

8. Implement the most preferred combination of the factors and their

levels in the design of a new product.

Application:

The two factor approach is useful when a manager goes for market
segmentation to promote his product. The approach will enable the top
level management to evolve a policy decision as to which segment of the
market has to be concentrated more in order to maximize the profit from
the product under consideration.

SELF ASSESSMENT QUESTIONS---END OF CHAPTER QUES-

TIONS[IMPORTANT FOR EXAMS]

1. Explain the purpose of ‘factor analysis’.

2. What is the objective of ‘conjoint analysis’? Explain.
3. State the steps in the development of conjoint analysis.
4. State the applications of conjoint analysis.
208
5. Enumerate the advantages and disadvantages of conjoint analysis.
6. What is a ‘product profile’? Explain.
7. What are the steps in multi-factor evaluation approach in conjoint
analysis?
8. What is a ‘two-factor table’? Explain.
9. Explain two-factor evaluation approach in conjoint analysis.

209
 Statistical Table-1: F-values at 1% level of significance

df1: degrees of freedom for greater variance

df2: degrees of freedom for smaller variance

df2
1 2 3 4 5 6 7 8 9 10
/df1
1 4052.1 4999.5 5403.3 5624.5 5763.6 5858.9 5928.3 5981.0 6022.4 6055.8

2 98.5 99.0 99.1 99.2 99.2 99.3 99.3 99.3 99.3 99.3
3 34.1 30.8 29.4 28.7 28.2 27.9 27.6 27.4 27.3 27.2
4 21.1 18.0 16.6 15.9 15.5 15.2 14.9 14.7 14.6 14.5
5 16.2 13.2 12.0 11.3 10.9 10.6 10.4 10.2 10.1 10.0
6 13.7 10.9 9.7 9.1 8.7 8.4 8.2 8.1 7.9 7.8
7 12.2 9.5 8.4 7.8 7.4 7.1 6.9 6.8 6.7 6.6
8 11.2 8.6 7.5 7.0 6.6 6.3 6.1 6.0 5.9 5.8
9 10.5 8.0 6.9 6.4 6.0 5.8 5.6 5.4 5.3 5.2
10 10.0 7.5 6.5 5.9 5.6 5.3 5.2 5.0 4.9 4.8
11 9.6 7.2 6.2 5.6 5.3 5.0 4.8 4.7 4.6 4.5
12 9.3 6.9 5.9 5.4 5.0 4.8 4.6 4.4 4.3 4.2
13 9.0 6.7 5.7 5.2 4.8 4.6 4.4 4.3 4.1 4.1
14 8.8 6.5 5.5 5.0 4.6 4.4 4.2 4.1 4.0 3.9
15 8.6 6.3 5.4 4.8 4.5 4.3 4.1 4.0 3.8 3.8
16 8.5 6.2 5.2 4.7 4.4 4.2 4.0 3.8 3.7 3.6
17 8.4 6.1 5.1 4.6 4.3 4.1 3.9 3.7 3.6 3.5
18 8.2 6.0 5.0 4.5 4.2 4.0 3.8 3.7 3.5 3.5
19 8.1 5.9 5.0 4.5 4.1 3.9 3.7 3.6 3.5 3.4
20 8.0 5.8 4.9 4.4 4.1 3.8 3.6 3.5 3.4 3.3
21 8.0 5.7 4.8 4.3 4.0 3.8 3.6 3.5 3.3 3.3
22 7.9 5.7 4.8 4.3 3.9 3.7 3.5 3.4 3.3 3.2
23 7.8 5.6 4.7 4.2 3.9 3.7 3.5 3.4 3.2 3.2
24 7.8 5.6 4.7 4.2 3.8 3.6 3.4 3.3 3.2 3.1
25 7.7 5.5 4.6 4.1 3.8 3.6 3.4 3.3 3.2 3.1
26 7.7 5.5 4.6 4.1 3.8 3.5 3.4 3.2 3.1 3.0
27 7.6 5.4 4.6 4.1 3.7 3.5 3.3 3.2 3.1 3.0
28 7.6 5.4 4.5 4.0 3.7 3.5 3.3 3.2 3.1 3.0
29 7.5 5.4 4.5 4.0 3.7 3.4 3.3 3.1 3.0 3.0
30 7.5 5.3 4.5 4.0 3.6 3.4 3.3 3.1 3.0 2.9

210
 Statistical Table-2: F-values at 2.5% level of significance

df1: degrees of freedom for greater variance

df2: degrees of freedom for smaller variance

df2
1 2 3 4 5 6 7 8 9 10
/df1
1 647.7 799.5 864.1 899.5 921.8 937.1 948.2 956.6 963.2 968.6
2 38.5 39.0 39.1 39.2 39.2 39.3 39.3 39.3 39.3 39.3
3 17.4 16.0 15.4 15.1 14.8 14.7 14.6 14.5 14.4 14.4
4 12.2 10.6 9.9 9.6 9.3 9.1 9.0 8.9 8.9 8.8
5 10.0 8.4 7.7 7.3 7.1 6.9 6.8 6.7 6.6 6.6
6 8.8 7.2 6.5 6.2 5.9 5.8 5.6 5.5 5.5 5.4
7 8.0 6.5 5.8 5.5 5.2 5.1 4.9 4.8 4.8 4.7
8 7.5 6.0 5.4 5.0 4.8 4.6 4.5 4.4 4.3 4.2
9 7.2 5.7 5.0 4.7 4.4 4.3 4.1 4.1 4.0 3.9
10 6.9 5.4 4.8 4.4 4.2 4.0 3.9 3.8 3.7 3.7
11 6.7 5.2 4.6 4.2 4.0 3.8 3.7 3.6 3.5 3.5
12 6.5 5.0 4.4 4.1 3.8 3.7 3.6 3.5 3.4 3.3
13 6.4 4.9 4.3 3.9 3.7 3.6 3.4 3.3 3.3 3.2
14 6.2 4.8 4.2 3.8 3.6 3.5 3.3 3.2 3.2 3.1
15 6.1 4.7 4.1 3.8 3.5 3.4 3.2 3.1 3.1 3.0
16 6.1 4.6 4.0 3.7 3.5 3.3 3.2 3.1 3.0 2.9
17 6.0 4.6 4.0 3.6 3.4 3.2 3.1 3.0 2.9 2.9
18 5.9 4.5 3.9 3.6 3.3 3.2 3.0 3.0 2.9 2.8
19 5.9 4.5 3.9 3.5 3.3 3.1 3.0 2.9 2.8 2.8
20 5.8 4.4 3.8 3.5 3.2 3.1 3.0 2.9 2.8 2.7
21 5.8 4.4 3.8 3.4 3.2 3.0 2.9 2.8 2.7 2.7
22 5.7 4.3 3.7 3.4 3.2 3.0 2.9 2.8 2.7 2.7
23 5.7 4.3 3.7 3.4 3.1 3.0 2.9 2.8 2.7 2.6
24 5.7 4.3 3.7 3.3 3.1 2.9 2.8 2.7 2.7 2.6
25 5.6 4.2 3.6 3.3 3.1 2.9 2.8 2.7 2.6 2.6
26 5.6 4.2 3.6 3.3 3.1 2.9 2.8 2.7 2.6 2.5
27 5.6 4.2 3.6 3.3 3.0 2.9 2.8 2.7 2.6 2.5
28 5.6 4.2 3.6 3.2 3.0 2.9 2.7 2.6 2.6 2.5
29 5.5 4.2 3.6 3.2 3.0 2.8 2.7 2.6 2.5 2.5
30 5.5 4.1 3.5 3.2 3.0 2.8 2.7 2.6 2.5 2.5

211
 Statistical Table-3: F-values at 5% level of significance

df1: degrees of freedom for greater variance

df2: degrees of freedom for smaller variance

df2
1 2 3 4 5 6 7 8 9 10
/df1
1 161.4 199.5 215.7 224.5 230.1 233.9 236.7 238.8 240.5 241.8
2 18.5 19.0 19.1 19.2 19.2 19.3 19.3 19.3 19.3 19.3
3 10.1 9.5 9.2 9.1 9.0 8.9 8.8 8.8 8.8 8.7
4 7.7 6.9 6.5 6.3 6.2 6.1 6.0 6.0 5.9 5.9
5 6.6 5.7 5.4 5.1 5.0 4.9 4.8 4.8 4.7 4.7
6 5.9 5.1 4.7 4.5 4.3 4.2 4.2 4.1 4.0 4.0
7 5.5 4.7 4.3 4.1 3.9 3.8 3.7 3.7 3.6 3.6
8 5.3 4.4 4.0 3.8 3.6 3.5 3.5 3.4 3.3 3.3
9 5.1 4.2 3.8 3.6 3.4 3.3 3.2 3.2 3.1 3.1
10 4.9 4.1 3.7 3.4 3.3 3.2 3.1 3.0 3.0 2.9
11 4.8 3.9 3.5 3.3 3.2 3.0 3.0 2.9 2.8 2.8
12 4.7 3.8 3.4 3.2 3.1 2.9 2.9 2.8 2.7 2.7
13 4.6 3.8 3.4 3.1 3.0 2.9 2.8 2.7 2.7 2.6
14 4.6 3.7 3.3 3.1 2.9 2.8 2.7 2.6 2.6 2.6
15 4.5 3.6 3.2 3.0 2.9 2.7 2.7 2.6 2.5 2.5
16 4.4 3.6 3.2 3.0 2.8 2.7 2.6 2.5 2.5 2.4
17 4.4 3.5 3.1 2.9 2.8 2.6 2.6 2.5 2.4 2.4
18 4.4 3.5 3.1 2.9 2.7 2.6 2.5 2.5 2.4 2.4
19 4.3 3.5 3.1 2.8 2.7 2.6 2.5 2.4 2.4 2.3
20 4.3 3.4 3.0 2.8 2.7 2.5 2.5 2.4 2.3 2.3
21 4.3 3.4 3.0 2.8 2.6 2.5 2.4 2.4 2.3 2.3
22 4.3 3.4 3.0 2.8 2.6 2.5 2.4 2.4 2.3 2.3
23 4.2 3.4 3.0 2.7 2.6 2.5 2.4 2.3 2.3 2.2
24 4.2 3.4 3.0 2.7 2.6 2.5 2.4 2.3 2.3 2.2
25 4.2 3.3 2.9 2.7 2.6 2.4 2.4 2.3 2.2 2.2
26 4.2 3.3 2.9 2.7 2.5 2.4 2.3 2.3 2.2 2.2
27 4.2 3.3 2.9 2.7 2.5 2.4 2.3 2.3 2.2 2.2
28 4.1 3.3 2.9 2.7 2.5 2.4 2.3 2.2 2.2 2.1
29 4.1 3.3 2.9 2.7 2.5 2.4 2.3 2.2 2.2 2.1
30 4.1 3.3 2.9 2.6 2.5 2.4 2.3 2.2 2.2 2.1

212
 Statistical Table-4: F-values at 10% level of significance

df1: degrees of freedom for greater variance

df2: degrees of freedom for smaller variance

df2/
1 2 3 4 5 6 7 8 9 10
df1
1 39.8 49.5 53.5 55.8 57.2 58.2 58.9 59.4 59.8 60.1
2 8.5 9.0 9.1 9.2 9.2 9.3 9.3 9.3 9.3 9.3
3 5.5 5.4 5.3 5.3 5.3 5.2 5.2 5.2 5.2 5.2
4 4.5 4.3 4.1 4.1 4.0 4.0 3.9 3.9 3.9 3.9
5 4.0 3.7 3.6 3.5 3.4 3.4 3.3 3.3 3.3 3.2
6 3.7 3.4 3.2 3.1 3.1 3.0 3.0 2.9 2.9 2.9
7 3.5 3.2 3.0 2.9 2.8 2.8 2.7 2.7 2.7 2.7
8 3.4 3.1 2.9 2.8 2.7 2.6 2.6 2.5 2.5 2.5
9 3.3 3.0 2.8 2.6 2.6 2.5 2.5 2.4 2.4 2.4
10 3.2 2.9 2.7 2.6 2.5 2.4 2.4 2.3 2.3 2.3
11 3.2 2.8 2.6 2.5 2.4 2.3 2.3 2.3 2.2 2.2
12 3.1 2.8 2.6 2.4 2.3 2.3 2.2 2.2 2.2 2.1
13 3.1 2.7 2.5 2.4 2.3 2.2 2.2 2.1 2.1 2.1
14 3.1 2.7 2.5 2.3 2.3 2.2 2.1 2.1 2.1 2.0
15 3.0 2.6 2.4 2.3 2.2 2.2 2.1 2.1 2.0 2.0
16 3.0 2.6 2.4 2.3 2.2 2.1 2.1 2.0 2.0 2.0
17 3.0 2.6 2.4 2.3 2.2 2.1 2.1 2.0 2.0 2.0
18 3.0 2.6 2.4 2.2 2.1 2.1 2.0 2.0 2.0 1.9
19 2.9 2.6 2.3 2.2 2.1 2.1 2.0 2.0 1.9 1.9
20 2.9 2.5 2.3 2.2 2.1 2.0 2.0 1.9 1.9 1.9
21 2.9 2.5 2.3 2.2 2.1 2.0 2.0 1.9 1.9 1.9
22 2.9 2.5 2.3 2.2 2.1 2.0 2.0 1.9 1.9 1.9
23 2.9 2.5 2.3 2.2 2.1 2.0 1.9 1.9 1.9 1.8
24 2.9 2.5 2.3 2.1 2.1 2.0 1.9 1.9 1.9 1.8
25 2.9 2.5 2.3 2.1 2.0 2.0 1.9 1.9 1.8 1.8
26 2.9 2.5 2.3 2.1 2.0 2.0 1.9 1.9 1.8 1.8
27 2.9 2.5 2.2 2.1 2.0 2.0 1.9 1.9 1.8 1.8
28 2.8 2.5 2.2 2.1 2.0 1.9 1.9 1.9 1.8 1.8
29 2.8 2.4 2.2 2.1 2.0 1.9 1.9 1.8 1.8 1.8
30 2.8 2.4 2.2 2.1 2.0 1.9 1.9 1.8 1.8 1.8

***

213
214
 CHAPTER V

1. Structure And Components Of Research Reports

What Is A Report?
Characteristics Of A Good Report
Framework Of A Report
Practical Reports Vs Academic Reports
Parts Of A Research Report
A Note On Literature Review

Understand the meaning of a research report

Analyze the components of a good report
Structure of a report
Characteristic differences in research report-
ing

WHAT IS A REPORT?

A report is a written document on a particular topic, which conveys

information and ideas and may also make recommendations. Reports
often form the basis of crucial decision making. Inaccurate, incomplete
and poorly written reports fail to achieve their purpose and reflect on the
decision, which will ultimately be made. This will also be the case if the
report is excessively long, jargonistic and/ or structureless. A good report
can be written by keeping the following features in mind:

1. All points in the report should be clear to the intended reader.

2. The report should be concise with information kept to a necessary

215
 minimum and arranged logically under various headings and sub-
headings.

3. All information should be correct and supported by evidence.

4. All relevant material should be included in a complete report.

Purpose Of Research Report

1. Why am i writing this report? Do i want to inform/ explain/ persuade,

or indeed all of these.

2. Who is going to read this report? Managers/ academicians/

researchers! What do they already know? What do they need to
know? Do any of them have certain attitudes or prejudices?

3. What resources do we have? Do i have access to a computer? Do i

have enough time? Can any of my colleagues help?

4. Think about the content of your report – what am i going to put in it?
What are my main themes? How much should be the text, and how
much should be the illustrations?

Framework Of A Report

The various frameworks can be used depending on the content

of the report, but generally the same rules apply. Introduction, method,
results and discussion with references or bibliography at the end, and an
abstract at the beginning could form the framework.

STRUCTURE OF A REPORT

Structure your writing around the IMR&D framework and you will
ensure a beginning, middle and end to your report.

I Introduction Why did i do this research? (beginning)

What did i do and how did i

M Method (middle)
go about doing it?
R Results What did i find? (middle)
AND

216
 D Discussion What does it all mean? (end)

What do I put in the beginning part?

Title of project, Sub–title (where

TITLE PAGE appropriate), Date, Author,
Organization, Logo
BACKGROUND History(if any) behind project
Author thanks people and
ACKNOWLEDGEMENT organization who helped during
the project
A condensed version of a
report – outlines salient points,
emphasizes main conclusions
SUMMARY(sometimes
and (where appropriate) the main
called abstract of the synopsis)
recommendations. N.B this is often
difficult to write and it is suggested
that you write it last.
An at- a – glance list that tells the
LIST OF CONTENTS reader what is in the report and
what page number(s) to find it on.
LIST OF TABLES As above, specifically for tables.
As above, specifically for
LIST OF APPENDICES
appendices.
Author sets the scene and states
INTRODUCTION
his/ her intentions.
AIMS – general aims of the
audit/ project, broad statement of
AIMS AND OBJECTIVES intent. OBJECTIVES – specific
things expected to do/ deliver(e.g.
expected outcomes)

217
What do I Put In the middle Part?

Work steps; what was done – how,

METHOD
by whom, when?
Honest presentation of the findings,
whether these were as expected
RESULT/FINDINGS or not. Give the facts, including
any inconsistencies or difficulties
encountered

What do I put in the end part?

Explanation of the results.( you

might like to keep the SWOT
analysis in mind and think about
DISCUSSION
your project’s strengths, weakness,
opportunities and threats, as you
write)
The author links the results/
findings with the points made in the
introduction and strives to reach
clear, simply stated and unbiased
CONCLUSIONS
conclusions. Make sure they are
fully supported by evidence and
arguments of the main body of
your audit/project.
The author states what specific
actions should be taken, by whom
and why. They must always be
RECOMMENDATIONS
linked to the future and should
always be realistic. Don’t make
them unless asked to.
A section of a report, which
provides full details of publications
REFERENCES
mentioned in the text, or from
which extracts have been quoted.

218
 The purpose of an appendix is
to supplement the information
APPENDIX
contained in the main body of the
report.

PRACTICAL REPORTS VS. ACADEMIC REPORTS

Practical Reports:

In the practical world of business or government, a report conveys

an information and (sometimes) recommendations from a researcher
who has investigated a topic in detail. A report like this will usually
be requested by people who need the information for a specific
purpose and their request may be written in terms of reference or
the brief. Whatever the report, it is important to look at the instruction
for what is wanted. A report like this differs from an essay in that
it is designed to provide information which will be acted on, rather
than to be read by people interested in the ideas for their own sake.
Because of this, it has a different structure and layout.

Academic Reports:

A report written for an academic course can be thought of

as a simulation. We can imagine that someone wants the report
for a practical purpose, although we are really writing the report as an
academic exercise for assessment. Theoretical ideas will be more to the
front in an academic report than in a practical one. Sometimes a report
seems to serve academic and practical purposes. Students on placement
with organizations often have to produce a report for the organization and
for assessment on the course. Although the background work for both
will be related, in practice, the report the student produces for academic
assessment will be different from the report produced for the organization,
because the needs of each are different.

219
RESEARCH REPORT: PRELIMINARIES

It is not sensible to leave all your writing until the end. There is
always the possibility that it will take much longer than you anticipate and
you will not have enough time. There could also be pressure upon available
word processors as other students try to complete their own reports. It is
wise to begin writing up some aspects of your research as you go along.
Remember that you do not have to write your report in the order it will
be read. Often it is easiest to start with the method section. Leave the
introduction and the abstract to last. The use of a word processor makes
it very straightforward to modify and rearrange what you have written
as your research progresses and your ideas change. The very process of
writing will help your ideas to develop. Last but by no means least, ask
someone to proofread your work.

STRUCTURE OF A RESEARCH REPORT

A research report has a different structure and layout in comparison

to a project report. A research report is for reference and is often quite a
long document. It has to be clearly structured for the readers to quickly
find the information wanted. It needs to be planned carefully to make
sure that the information given in the report is put under correct
headings.

PARTS OF RESEARCH REPORT

Cover sheet: This should contain some or all of the following:

Full title of the report
Name of the researcher
Name of the unit of which the project is a part
Name of the institution
Date/year.

Title page: Full title of the report.

Your name
Acknowledgement: a thanks giving to the people who helped you.
Contents

220
List of the Tables

Headings and sub-headings used in the report should be given

with their page numbers. Each chapter should begin on a new page. Use
a consistent system in dividing the report into parts. The simplest may be
to use chapters for each major part and subdivide these into sections and
sub-sections. 1, 2, 3 etc. Can be used as the numbers for each chapter. The
sections of chapter 3 (for example) would be 3.1, 3.2, 3.3, and so on. For
further sub-division of a sub-section you may use 3.2.1, 3.2.2, and so on.

Abstract or Summary or Executive Summary or Introduction:

This presents an overview of the whole report. It should let the

reader see in advance, what is in the report. This includes what you set out
to do, how review of literature is focused and narrowed in your research,
the relation of the methodology you chose to your objectives, a summary
of your findings and analysis of the findings

BODY

Aims And Purpose or Aims And Objectives:

Why did you do this work? What was the problem you were
investigating? If you are not including review of literature, mention the
specific research/es which is/are relevant to your work.

Review of Literature

This should help to put your research into a background context

and to explain its importance. Include only the books and articles which
relate directly to your topic. You need to be analytical and critical, and not
just describe the works that you have read.

Methodology

Methodology deals with the methods and principles used in an activity,

in this case research. In the methodology chapter, explain the method/s
you used for the research and why you thought they were the appropriate
221
ones. You may, for example, be depending mostly upon secondary data or
you might have collected your own data. You should explain the method of
data collection, materials used, subjects interviewed, or places you visited.
Give a detailed account of how and when you carried out your research and
explain why you used the particular method/s, rather than other methods.
Included in this chapter should be an examination of ethical issues, if any.

Results or Findings

What did you find out? Give a clear presentation of your results.
Show the essential data and calculations here. You may use tables, graphs
and figures.

Analysis and Discussion

Interpret your results. What do you make out of them? How do

they compare with those of others who have done research in this area?
The accuracy of your measurements/results should be discussed and
deficiencies, if any, in the research design should be mentioned.

Conclusions

What do you conclude? Summarize briefly the main conclusions

which you discussed under “Results.” were you able to answer some or all
of the questions which you raised in your aims and objectives? Do not be
tempted to draw conclusions which are not backed up by your evidence.
Note the deviation/s from expected results and any failure to achieve all
that you had hoped.

Recommendations

Make your recommendations, if required. The suggestions for

action and further research should be given.

Appendix

You may not need an appendix, or you may need several. If you have
used questionnaires, it is usual to include a blank copy in the appendix.
You could include data or calculations, not given in the body, that are
222
necessary, or useful, to get the full benefit from your report. There may
be maps, drawings, photographs or plans that you want to include. If you
have used special equipment, you may include information about it.

The plural of an appendix is appendices. If an appendix or

appendices are needed, design them thoughtfully in a way that your
readers find it/them convenient to use.

References

List all the sources which you referred in the body of the report. You
may use the pattern prescribed by American Psychological Association, or
any other standard pattern recognized internationally.

REVIEW OF LITERATURE

In the case of small projects, this may not be in the form of a

critical review of the literature, but this is often asked for and is a standard
part of larger projects. Sometimes students are asked to write Review of
Literature on a topic as a piece of work in its own right. In its simplest
form, the review of literature is a list of relevant books and other sources,
each followed by a description and comment on its relevance.

The literature review should demonstrate that you have read and
analysed the literature relevant to your topic. From your readings, you
may get ideas about methods of data collection and analysis. If the review
is part of a project, you will be required to relate your readings to the issues
in the project, and while describing the readings, you should apply them
to your topic. A review should include only relevant studies. The review
should provide the reader with a picture of the state of knowledge in the
subject.
Your literature search should establish what previous researches
have been carried out in the subject area. Broadly speaking, there are three
kinds of sources that you should consult:

1. Introductory material;
2. Journal articles and
3. Books.
To get an idea about the background of your topic, you may consult
223
one or more textbooks at the appropriate time. It is a good practice to
review in cumulative stages - that is, do not think you can do it all at
one go. Keep a careful record of what you have searched, how you have
gone about it, and the exact citations and page numbers of your readings.
Write notes as you go along. Record suitable notes on everything you read
and note the methods of investigations. Make sure that you keep a full
reference, complete with page numbers. You will have to find your own
balance between taking notes that are too long and detailed, and ones too
brief to be of any use. It is best to write your notes in complete sentences
and paragraphs, because research has shown that you are more likely to
understand your notes later if they are written in a way that other people
would understand. Keep your notes from different sources and/or about
different points on separate index cards or on separate sheets of paper.
You will do mainly basic reading while you are trying to decide on your
topic. You may scan and make notes on the abstracts or summaries of
work in the area. Then do a more thorough job of reading later on, when
you are more confident of what you are doing. If your project spans several
months, it would be advisable towards the end to check whether there are
any new and recent references.

REFERENCES

There are many methods of referencing your work; some of the

most common ones are the numbered style, american psychological
association style and the harvard method, with many other variations. Just
use the one you are most familiar and comfortable with. Details of all the
works referred by you should be given in the reference section.

THE PRESENTATION OF REPORT

Well-produced, appropriate illustrations enhance the presentability

of a report. With today’s computer packages, almost anything is possible.
However, histograms, bar charts and pie charts are still the three ‘staples’.
Readers like illustrated information, because it is easier to absorb and
it’s more memorable. Illustrations are useful only when they are easier to
understand than words or figures and they must be relevant to the text.
Use the algorithm included to help you decide whether or not to use an
illustration. They should never be included for their own sake, and don’t
overdo it; too many illustrations distract the attention of readers.
224
 CHAPTER V

2. Types Of Reports: Characteristics Of Good

Research Report

Different Types Of Reports

Technical Reports
General Reports
Reporting Styles
Characteristics Of A Good Re-
port

Understand different types of reports

Technical reports and their contents
General reports
Different types of writing styles
Essential characteristics of a good re-
port

Reports vary in length and type. Students’ study reports are often
called term papers, project reports, theses, dissertations depending on the
nature of the report. Reports of researchers are in the form of monographs,
research papers, research thesis, etc. In business organizations a wide
variety of reports are under use: project reports, annual reports of financial
statements, report of consulting groups, project proposals etc. News items
in daily papers are also one form of report writing. In this lesson, let us
identify different forms of reports and their major components.

225
Types of reports

Reports may be categorized broadly as Technical Reports and

General Reports based on the nature of methods, terms of reference and
the extent of in-depth enquiry made etc. On the basis of usage pattern,
the reports may also be classified as information oriented reports, decision
oriented reports and research based reports. Further, reports may also
differ based on the communication situation. For example, the reports may
be in the form of Memo, which is appropriate for informal situations or for
short periods. On the other hand, the projects that extend over a period of
time, often call for project reports. Thus, there is no standard format of
reports. The most important thing that helps in classifying the reports is
the outline of its purpose and answers for the following questions:

ӹӹ What did you do?

ӹӹ Why did you choose the particular research method that you
used?
ӹӹ What did you learn and what are the implications of what you
learned?
ӹӹ If you are writing a recommendation report, what action are you
recommending in response to what you learned?

Two types of report formats are described below:

A Technical Report

A technical report mainly focuses on methods employed,

assumptions made while conducting a study, detailed presentation of
findings and drawing inferences and comparisons with earlier findings
based on the type of data drawn from the empirical work.
An outline of a Technical Report mostly consists of the following:
Title and nature of the study:

Brief title and the nature of work sometimes followed by subtitle

indicate more appropriately either the method or tools used. Description
of objectives of the study, research design, operational terms, working
hypothesis, type of analysis and data required should be present.

226
Abstract of Findings:

A brief review of the main findings just can be made either in a
paragraph or in one/two pages.
Review of current status:

A quick review of past observations and contradictions reported,
applications observed and reported are reviewed based on the in-house
resources or based on published observations.
Sampling and Methods employed

Specific methods used in the study and their limitations. In the case
of experimental methods, the nature of subjects and control conditions are
to be specified. In the case of sample studies, details of the sample design
i.e., sample size, sample selection etc are given.

Data sources and experiment conducted

Sources of data, their characteristics and limitations should be

specified. In the case of primary survey, the manner in which data has
been collected should be described.

Analysis of data and tools used.

The analysis of data and presentation of findings of the study with
supporting data in the form of tables and charts are to be narrated. This
constitutes the major component of the research report.

Summary of findings

A detailed summary of findings of the study and major observations
should be stated. Decision inputs if any, policy implications from the
observations should be specified.
References
A brief list of studies conducted on similar lines, either preceding the
present study or conducted under different experimental conditions is
listed.

227
Technical appendices

These appendices include the design of experiments or
questionnaires used in conducting the study, mathematical derivations,
elaboration on particular techniques of analysis etc.
General Reports

General reports often relate popular policy issues mostly related

to social issues. These reports are generally simple, less technical, good
use of tables and charts. Most often they reflect the journalistic style.
Example for this type of report is the “Best B-Schools Survey in Business
Magazines”. The outline of these reports is as follows:

1. Major Findings and their Implications

2. Recommendations for Action

3. Objectives of the Study

4. Method Employed for Collecting Data

5. Results

Writing Styles

There are atleast 3 distinct report writing styles that can be applied
by students of Business Studies. They are called:

1. Conservative
2. Key points
3. Holistic

1. Conservative Style

Essentially, the conservative approach takes the best structural

elements from essay writing and integrates these with appropriate report
writing tools. Thus, headings are used to deliberate upon different sections
of the answer. In addition, the space is well utilized by ensuring that each
paragraph is distinct (perhaps separated from other paragraphs by leaving
two blank lines in between).

228
2. Key Point Style

This style utilizes all of the report writing tools and is thus more
overtly ‘report-looking’. Use of headings, underlining, margins, diagrams
and tables are common. Occasionally reporting might even use indentation
and dot points. The important thing to remember is that the tools should
be applied in a way that adds to the report. The question must be addressed
and the tools applied should assist in doing that. An advantage of this style
is the enormous amount of information that can be delivered relatively
quickly.

3. Holistic Style

The most complex and unusual of the styles, holistic report writing
aims to answer the question from a thematic and integrative perspective.
This style of report writing requires the researcher to have a strong
understanding of the course and is able to see which outcomes are being
targeted by the question.

Essentials Of A Good Report:

Good research report should satisfy some of the following basic

characteristics:

STYLE

Reports should be easy to read and understand. The style of the

writer should ensure that sentences are succinct and the language used is
simple, to the point and avoiding excessive jargon.

LAYOUT

A good layout enables the reader to follow the report’s intentions,

and aids the communication process. Sections and paragraphs should be
given headings and sub¬-headings. You may also consider a system of
numbering or lettering to identify the relative importance of paragraphs
and sub-paragraphs. Bullet points are an option for highlighting important
points in your report.

229
ACCURACY

Make sure everything you write is factually accurate. If you would

mislead or misinform, you will be doing a disservice not only to yourself
but also to the readers, and your credibility will be destroyed. Remember
to refer to any information you have used to support your work.

CLARITY

Take a break from writing. When you would come back to it, you’ll
have the degree of objectivity that you need. Use simple language to express
your point of view.

READABILITY

Experts agree that the factors, which affect readability the most, are:

ӹӹ Attractive appearance
ӹӹ Non-technical subject matter
ӹӹ Clear and direct style
ӹӹ Short sentences
ӹӹ Short and familiar words

REVISION

When first draft of the report is completed, it should be put to one

side atleast for 24 hours. The report should then be read as if with eyes of
the intended reader. It should be checked for spelling and grammatical
errors. Remember the spell and grammar check on your computer. Use it!

REINFORCEMENT

Reinforcement usually gets the message across. This old adage is
well known and is used to good effect in all sorts of circumstances e.g.,
presentations - not just report writing.

ӹӹ TELL THEM WHAT YOU ARE GOING TO SAY: in the

introduction and summary you set the scene for what follows in
your report.
230
 ӹӹ THEN SAY IT : you spell things out in results/findings

ӹӹ THEN TELL THEM WHAT YOU SAID: you remind your readers
through the discussion what it was all about.

FEEDBACK MEETING

It is useful to circulate copies of your report prior to the feedback

meeting. Meaningful discussion can then take place during the feedback
meeting with recommendations for change more likely to be agreed upon
which can then be included in your conclusion. The following questions
should be asked at this stage to check whether the Report served the
purpose:

ӹӹ Does the report have impact?

ӹӹ Do the summary /abstract do justice to the report?
ӹӹ Does the introduction encourage the reader to read more?
ӹӹ Is the content consistent with the purpose of the report?
ӹӹ Have the objectives been met?
ӹӹ Is the structure logical and clear?
ӹӹ Have the conclusions been clearly stated?
ӹӹ Are the recommendations based on the conclusions and expressed
clearly and logically?

***

231
232
 CHAPTER V

3. Format And Presentation Of A Report

Importance Of Presentation Of A Report

Common Elements Of A Format
Title Page
Introductory Pages
Body Of The Text
References
Appendix
Dos And Don’ts
Presentation Of Reports

Understand the importance of format of a report

Contents of a title page
What should be in introductory pages
Contents of a body text
How to report other studies
Contents of an appendix
Dos and don’ts of a report

Any report serves its purpose, if it is finally presented before the

stakeholders of the work. In the case of an MBA student, Project Work
undertaken in an industrial enterprise and the findings of the study would
be more relevant, if they are presented before the internal managers of
the company. In the case of reports prepared out of consultancy projects,

233
a presentation would help the users to interact with the research team
and get clarification on any issue of their interest. Business Reports or
Feasibility Reports do need a summary presentation, if they have to serve
the intended purpose. Finally, the Research Reports of the scholars would
help in achieving the intended academic purpose, if they are made public
in academic symposiums, seminars or in Public Viva Voce examinations.
Thus, the presentation of a report goes along with preparation of a good
report. Further, the use of graphs, charts, citations and pictures draw the
attention of readers and audience of any type. In this lesson, it is intended
to provide a general outline related to the presentation of any type of
report. See Exhibit I

Exhibit I

Common Elements Of A Report

A report may contain some or all of the following, please refer to

your departmental guidelines.

MEMORANDUM OR COVERING LETTER

Memorandum or Covering Letter is a brief note stating the purpose

or giving an explanation that is used when the report is sent to someone
within the same organization.

TITLE PAGE

It is addressed to the receiver of a report while giving an explanation

for it, and is used when the report is for someone who does not belong to
the same organization as the writer. It contains a descriptive heading or
name. It may also contain author’s name, position, company’s name and so
on.

EXECUTIVE SUMMARY

Executive summary summarizes the main contents and is usually

of about 300-350 words.

234
TABLE OF CONTENTS

Table of Contents consists of a list of the main sections, indicating

the page on which each section begins.

INTRODUCTION

Informs the reader of what the report is about—aim and purpose,

significant issues, any relevant background information.

REVIEW OF LITERATURE

Presents critical analysis of the available research to build a base for

the present study.

METHODOLOGY

Gives details about nature of the study, research design, sample,

and tools used for data collection and analysis.

RESULTS

Presents findings of the study.

DISCUSSION

Describes the reasoning and research in detail.

CONCLUSION/S

Summarizes the main points made in the written work in the

light of objectives. It often includes an overall answer to the problem/s
addressed; or an overall statement synthesizing the strands of information
dealt with.

RECOMMENDATION/S OR IMPLICATIONS

Gives suggestions related to the issue(s) or problem(s) dealt with. It may

highlight the applications of the findings under implications section.
235
REFERENCES

An alphabetical list of all sources referred in the report.

APPENDICES

Extra information of further details placed after the main body of

the text.

FORMATS OF REPORTS

Before attempting to look into presentation dimensions of a Report,

a quick look into standard format associated with a Research Report is
examined hereunder. The format generally includes the steps one should
follow while writing and finalizing their research report.

Different Parts of a Report

Generally different parts of a report include:

1. Cover Page / Title Page

2. Introductory Pages ( Foreword, Preface, Acknowledgement, Table of

Contents, List of Tables, List of Illustrations or Figures, Key Words /
Abbreviations Used Etc.)

3. Contents of the Report (Which Generally Includes a Macro Setting,

Research Problem, Methodology Used, Objectives of the Study,
Review of Studies, Tools Used for Data Collection and Analysis,
Empirical Results in One/Two Sections, Summary of Observations
etc.)

4. References (Including Appendices, Glossary of Terms Used, Source

Data, Derivations of Formulas for Models Used in the Analysis etc.)

236
Title Page:

The Cover Page or Title Page of a Research Report should contain

the following information:

1. Title of the Project / Subject

2. Who has conducted the study

3. For what purpose

4. Organization

5. Period of submission

***

237
238
A Model:

An example of a Summer Project Report conducted by an MBA

student generally follows the following Title Page:

A STUDY ON THE USE OF COMPUTER TECHNOLOGY IN BANKING

OPERATIONS IN XXX BANK LTD., PONDICHERRY

A SUMMER PROJECT REPORT

PREPARED BY
Ms. MADAVI LATHA

Submitted at

SHIVANI INSTITUTE OF
BUSINESS MANAGEMENT
MANGALORE-575003

Introductory Pages:

Introductory pages generally do not constitute the Write up of the

Research work done. These introductory pages basically form the Index
of the work done. These pages are usually numbered in Roman numerical
(eg, I, ii, iii etc). The introductory pages include the following components
239
 ӹӹ Foreword
ӹӹ Preface
ӹӹ Acknowledgements
ӹӹ Table of Contents
ӹӹ List of Tables
ӹӹ List of Figures / Charts

Foreword is usually one page write up or a citation about the

work by any eminent / popular personality or a specialist in the given
field of study. Generally, the write up includes a brief background on
the contemporary issues and suitability of the present subject and its
timeliness, major highlights of the present work, brief background of the
author etc. The writer of the foreword generally gives the foreword on his
letter head

Preface is again one/two pages write up by the author of the book

/ report stating circumstances under which the present work is taken
up, importance of the work, major dimensions examined and intended
audience for the given work. The author gives his signature and address at
the bottom of the page along with date and year of the work

Acknowledgements is a short section, mostly a paragraph. It

mostly consists of sentences giving thanks to all those associated and
encouraged to carry out the present work. Generally, author takes time
to acknowledge the liberal funding by any funding agency to carry out the
work, and agencies which had given permission to use their resources. At
the end, the author thanks everybody and gives his signature.

Table of Contents refers to the index of all pages of the said Research
Report. These contents provide the information about the chapters, sub-
sections, annexure for each chapter, if any, etc. Further, the page numbers
of the content of the report greatly helps any one to refer to those pages for
necessary details. Most authors use different forms while listing the sub
contents. These include alphabet classification and decimal classification.

Examples for both of them are given below:

Example of content sheet (alphabet classification)
An example of Content Sheet with decimal classification
240
 CONTENTS

Foreword i

Preface iii

Acknowledgement v

Chapter I (Title of the Chapter) INTRODUCTION

1. Macro economic background 1

2. Performance of a specific industry sector 6

3. Different studies conducted so far 9

4 nature and scope 17

4.1. Objectives of the study 18

4.2. Methodology adopted 19

4.2. A. Sampling Procedure adopted 20

4.2.b. Year of the study 20

Chapter II (Title of the Chapter): Empirical Results I 22

1. Test Results of H1 22

2. Test Results of H2 27

3 test Results of H3 32

3.1. Sub Hypothesis of H3 33

241
 3.2. Sub hypothesis of h2 37

Chapter III 45

Chapter IV 85

Chapter V (Summary & Conclusions) 120

Appendices 132

References/Bibliography 135

Glossary 140

List of Tables and Charts:

Details of Charts and Tables given in the Research Report are

numbered and presented on separate pages and the lists of such tables
and charts are given on a separate page. Tables are generally numbered
either in Arabic numerals or in decimal form. In the case of decimal form,
it is possible to indicate the chapter to which the said table belongs. For
example, Table 2.1 refers to Table 1 in Chapter 2.

Executive Summary:

Most Business Reports or Project works conducted on a specific

issue carry one or two pages of Executive Summary. This summary
precedes the chapters of the Regular Research Report. This summary
generally contains a brief description of problem under enquiry, methods
used and the findings. A line about the possible alternatives for decision
making would be the last line of the Executive Summary.

BODY OF THE REPORT:

The body of the Report is the most important part of the report.
This body of report may be segmented into a handful of Units or Chapters
242
arranged in a sequential order. Research Report often present the
Methodology, Objectives of the study, data tools, etc in the first or second
chapters along with a brief background of the study, review of relevant
studies. The major findings of the study are incorporated into two or three
chapters based on the major or minor hypothesis tested or based on the
sequence of bjectives of the study. Further, the chapter plan may also be
based likely on different dimensions of the problem under enquiry.

Each Chapter may be divided into sections. While the first section
may narrate the descriptive characteristics of the problem under enquiry,
the second and subsequent sections may focus on empirical results based
on deeper insights of the problem of study. Each chapter based on research
studies mostly contain major headings, sub headings, quotations drawn
from observations made by earlier writers, footnotes and exhibits.

Use of References:

There are two types of reference formatting. The first is the ‘in-text’
reference format, where previous researchers and authors are cited during
the building of arguments in the introduction and discussion sections.
The second type of format is that adopted for the Reference section for
writing footnotes or Bibliography.

Citations in the text

The names and dates of researchers go in the text as they are

mentioned e.g., “This idea has been explored in the work of Smith (1992).”
it is generally unacceptable to refer to authors and previous researchers
etc.

Examples of Citing References (Single Author)

Duranti (1995) has argued or it has been argued that (Duranti, 1995)

In the case of more authors,

Moore, Maguire, and Smyth (1992) proposed or it has been proposed

that (Moore, Macquire, & Smyth, 1992)

243
For subsequent citations in the same report: Moore et al.(1992) also
proposed... Or it has also been proposed that.... (M oore et al.,
1992)

The reference section:

The report ends with reference section, which comes immediately

after the Recommendations and begins on a new page. It is titled as
‘References’ in upper and lower case letters centered across the page.

Published Journal Articles

Beckerian, D.A. (1993). In search of the typical eyewitness.

American psychologist, 48, 574-576.

Gubbay, S.S., Ellis, W., Walton, J.N., and Court, S.D.M. (1965).
Clumsy Children: a study of apraxic and agnosic defects in 21 children.
Brain, 88, 295-312.

Authored Books
Cone, J.D., and Foster, S.L. (1993). Dissertations and theses from start
to finish: psychology and related fields. Washington, Dc: American
Psychological Association.

Cone, J.D., and Foster, S.L. (1993). Dissertations and theses from start to
finish: psychology and related fields (2nd Ed.). Washington, Dc: American
Psychological Association.

APPENDICES:

The purpose of the appendices is to supplement the main body of

your text and provide additional information that may be of interest to the
reader.

There is no major heading for the appendices. You simply need to

include each one, starting on a new page, numbered, using capital letters,
and headed with a centered brief descriptive title. For example:

Appendix A: List of stimulus words presented to the participants

244
Dos and Don’ts of Report Writing

1. Choose a font size that is not too small or too large; 11 or 12 is a good
font size to use.

2. Acknowledgment need not be a separate page, except in the final report.

In fact, you could just drop it altogether for the first- and second-stage
reports. Your guide already knows how much you appreciate his/her
support. Express your gratitude by working harder instead of writing a
flowery acknowledgment.

3. Make sure your paragraphs have some indentation and that it is not too
large. Refer to some text books or journal papers if you are not sure.

4. If figures, equations, or trends are taken from some reference, the

reference must be cited right there, even if you have cited it earlier.

5. The correct way of referring to a figure is Fig. 4 or Fig. 1.2 (note that
there is a space after Fig.). The same applies to Section, Equation, etc.
(e.g., sec. 2, eq. 3.1).

6. Cite a reference as, for example, “The threshold voltage is a strong

function of the implant dose [1].” note that there must be a space before
the bracket.

7. Follow some standard format while writing references. For example,

you could look up any IEEE transactions issue and check out the
format for journal papers, books, conference papers, etc.

8. Do not type references (for that matter, any titles or captions) entirely
in capital letters. The only capital letters required are (i) the first letter
of a name, (ii) acronyms, (iii) the first letter of the title of an article (iv)
the first letter of a sentence.

9. The order of references is very important. In the list of your references,

the first reference must be the one which is cited before any other
reference, and so on. Also, every reference in the list must be cited
at least once (this also applies to figures). In handling references and
figure numbers, Latex turns out to be far better than Word.

10. Many commercial packages allow “screen dump” of figures. While this
is useful in preparing reports, it is often very wasteful (in terms of
toner or ink) since the background is black. Please see if you can invert
245
 the image or use a plotting program with the raw data such that the
background is white.

11. The following tips may be useful: (a) for windows, open the file in
Paint and select Image/Invert Colors. (b) For Linux, open the file in
Image Magick (this can be done by typing display) and then selecting
Enhance/Negate.

12. As far as possible, place each figure close to the part of the text where
it is referred to.

13. A list of figures is not required except for the final project report. It
generally does not do more than wasting paper.

14. The figures, when viewed together with the caption, must be, as far as
possible, self-explanatory. There are times when one must say, “see text
for details”. However, this is an exception and not a rule.

15. The purpose of a figure caption is simply to state what is being

presented in the figure. It is not the right place for making comments
or comparisons; that should appear only in the text.

16. If you are showing comparison of two (or more) quantities, use the same
notation through out the report. For example, suppose you want to
compare measured data with analytical model in four different figures,
in each figure, make sure that the measured data is rep¬resented by
the same line type or symbol. The same should be followed for the
analytical model. This makes it easier for the reader to focus on the
important aspects of the report rather than getting lost in lines and
symbols.

17. If you must resize a plot or a figure, make sure that you do it
simultaneously in both x and y directions. Otherwise, circles in the
original figure will appear as ellipses, letters will appear too fat or too
narrow, and other similar calamities will occur.

18. In the beginning of any chapter, you need to add a brief introduction
and then start sections. The same is true about sections and subsections.
If you have sections that are too small, it only means that there is not
enough material to make a separate section. In that case, do not make
a separate section. Include the same material in the main section or
elsewhere.

246
19. Remember, a short report is perfectly acceptable if you have put in
the effort and covered all important aspects of your work. Adding
unnecessary sections and subsections will create the impression that
you are only covering up the lack of effort.

20. Do not make one-line paragraphs.

21. Always add a space after a full stop, comma, colon, etc. Also, leave a
space before opening a bracket. If the sentence ends with a closing
bracket, add the full stop (or comma or semicolon, etc) after the
bracket.

22. Do not add a space before a full stop, comma, colon, etc.

23. Using a hyphen can be tricky. If two (or more) words form a single
adjective, a hyphen is required; otherwise, it should not be used. For
example, (a) A short-channel device shows a finite output conductance.
(b) This is a good example of mixed-signal simulation. (c)Several
devices with short channels were studied.

24. If you are using Latex, do not use the quotation marks to open. If you

do that, you get “this”. Use the single opening quotes (twice) to get
“this”.

25. Do not use very informal language. Instead of “this theory should be
taken with a pinch of salt,” you might say, “this theory is not convincing,”
or “it needs more work to show that this theory applies in all cases.”

26. Do not use “&”; write “and” instead. Do not write “There’re” for “There
are” etc.

27. If you are describing several items of the same type (e.g., short-channel
effects in a MOS transistor), use the “list” option; it enhances the clarity
of your report.

28. Do not use “bullets” in your report. They are acceptable in a presentation,
but not in a formal report. You may use numerals or letters instead.

29. Whenever in doubt, look up a text book or a journal paper to verify

whether your grammar and punctuation are correct.

30. Do a spell check before you print out your document. It always helps.

31. Always write the report so that the reader can easily make out what
247
 your contribution is. Do not leave the reader guessing in this respect.
32. Above all, be clear. Your report must have a flow, i.e., the reader must
be able to appreciate continuity in the report. After the first reading,
the reader should be able to understand (a) the overall theme and (b)
what is new (if it is a project report).

33. Plagiarism is a very serious offense. You simply cannot copy material
from an existing report or paper and put it verbatim in your report.
The idea of writing a report is to convey in your words what you have
understood from the literature.

The above list may seem a little intimidating. However, if you make a
sincere effort, most of the points are easy to remember and practice.
A supplementary exercise that will help you immensely is that of
looking for all major and minor details when you read an article from a
newspaper or a magazine, such as grammar, punctuation, organization
of the material, etc.

PRESENTATION OF A REPORT

In this section, we will look into the issues associated with

presentation of a Research Report by the Researcher or Principal
Investigator. While preparing for the presentation of a report, the
researchers should focus on the following issues:

1. What is the purpose of the report and issues on which the

Presentation has to focus?

2. Who are the stakeholders and what are their areas of interest?

3. The mode and media of presentation.

4. Extent of Coverage and depth to address at.

5. Time, Place and cost associated with presentation.

6. Audio – Visual aids intended to be used.

248
Case Study

1. Satisfying Customer Need:

A customer is approaching a Godrej Refrigerator (Exclusively)
dealer. By considering the following how will you match the product to
the customer for the satisfaction?

a. Vegetarian/Non Vegetarian

b. Non Vegetarian/Fost Free

c. Size of Refrigerator and Family

d. Brand first/Shop First

e. Buying Behavior/Changing Behaviour

2. A leading software services company has lived to as a researcher.

The task before you is to assess the need for software products that will be
accepted readily in the market. The company has not looked at software
products and has no idea of the industry demand.Prepare a research
proposal to identify the industrial segment and the type of software
products that the company can consider manufacturing. The proposal
should include the research design, time line and limitations.

3. Given below are the figures of production (in lakh kg) of a sugar
factory.

Year: 1995 1996 1997 1998 1999 2000 2001

Production: 40 45 46 42 47 50 46

Fit a straight-line trend by the least squares method and tabulate the trend.

4. Genetic theory states that children having one parent of blood type
A and the other of blood type B will always be of one of three types, A,
AB, B and the proportion of three types will on an average be as 1 : 2 : 1.
A report states that out of 300 children having one A parent and B parent,
30 per cent were found to be types A, 45 per cent type AB and remainder
type B. Test the hypothesis by Chi square test.
***
249
250
 This book was distributed courtesy of:

For your own Unlimited Reading and FREE eBooks today, visit:
http://www.Free-eBooks.net

Share this eBook with anyone and everyone automatically by selecting any of the
options below:

To show your appreciation to the author and help others have

wonderful reading experiences and find helpful information too,
we'd be very grateful if you'd kindly
post your comments for this book here.

COPYRIGHT INFORMATION

Free-eBooks.net respects the intellectual property of others. When a book's copyright owner submits their work to Free-eBooks.net, they are granting us permission to distribute such material. Unless
otherwise stated in this book, this permission is not passed onto others. As such, redistributing this book without the copyright owner's permission can constitute copyright infringement. If you
believe that your work has been used in a manner that constitutes copyright infringement, please follow our Notice and Procedure for Making Claims of Copyright Infringement as seen in our Terms
of Service here:

http://www.free-ebooks.net/tos.html